is it overtraining? i looks like it was going up, but now its going down
#✨│ai-help
lime sorrel
warped hedge
thank you
molten pecan
is it overtraining? i looks like it was going up, but now its going down
Seems like it's hitting a new lowest point
So no, I don't think it's overtraining
blissful timber
bro can anyone help please
raven osprey
i got you
lime sorrel
Seems like it's hitting a new lowest point
this is what it look like with the "ignore outliner" button off
raven osprey
public as in upload it to here?
molten pecan
It might be overtraining now. Keep an eye if the line is going straight indefinitely
raven osprey
i put it in rvc testing
lime sorrel
It might be overtraining now. Keep an eye if the line is going straight indefini...
its now going up and down
i think its overtraining, do i take the lowest point?
desert niche
-colab
raven osprey
how do i fix wee not defined
storm rose
i've been having some problems with using RVC on Google Colab. when i go to download the model it keeps showing the following error:
"Cannot retrive the public link of the file", but I know that's not the issue cause i used these models until a couple days ago
does anyone know how i can fix this?
glad zealot
its now going up and down
Just keep going
It seems to still be going down so it's fine
lime sorrel
@glad zealot i think this should be enough
glad zealot
Still going down but you can get the model and test it out while still training
lime sorrel
i set the epochs at 300 and it started overtraining at around 3k, which means i kept going until i reached 6k
you sure its going down?
glad zealot
Yup
lime sorrel
im pretty sure it kept going up and down
rare gobletBOT
Ayo? @lime sorrel level 3 !!! 
lime sorrel
damn i stopped it cuz i thought it was overtraining
glad zealot
I mean it seems to already be good
glad zealot
I mean this is what I have and it sounds amazing so don't depend too much on the graph
lime sorrel
I mean this is what I have and it sounds amazing so don't depend too much on the...
ok
do i click the "skip models" for best results?
glad zealot
No clue
noble wyvern
PLEASE HELP i cant hear anything
rare gobletBOT
Ayo? @noble wyvern level 2 !!!
noble wyvern
after i use VAC
viscid moss
Put the same name as the first time you trained the model
viscid moss
index -1 is out of bounds for axis 0 with size 0 when i try to generate
HF rigth?
slate lagoon
?
viscid moss
Is that Hugging Face?
viscid moss
i used a different hf model and it works fine is it just a bad model
hmmmm I think the model has the wrong index
send me that model
slate lagoon
ok
viscid moss
lemme test
viscid moss
https://huggingface.co/NotSharpe/KidCudiRVC2/resolve/main/KidCudiRVC2.zip
wrong index file
brittle wing
does anyone know how to go back to previous versions of this google colab?
https://colab.research.google.com/drive/1r4IRL0UA7JEoZ0ZK8PKfMyTIBHKpyhcw
it changed and I'm really confused on how to use it
slate lagoon
oh
slate lagoon
will try different model
rare gobletBOT
Ayo? @slate lagoon level 3 !!!
slate lagoon
How did you get that model?
/search command
viscid moss
/search command
can u send me the link of the bot response pls
violet heron
does anyone know how to go back to previous versions of this google colab?
http...
It changed because the old one doesn’t work anymore
viscid moss
https://discord.com/channels/1159260121998827560/1163592055830880266/11956050342...
ty
brittle wing
It changed because the old one doesn’t work anymore
ah, okay sorry
slate lagoon
is there any way i can use that model
lime sorrel
I dont see the addex.index file, where is it?
the tutorial says its an important file
violet heron
I dont see the addex.index file, where is it?
Did you do the “train index”
lime sorrel
Did you do the “train index”
i forgot, thx
viscid moss
is there any way i can use that model
Use it without .index file. But I'm not sure if HF spaces can inference in that way
carmine osprey
Use it without .index file. But I'm not sure if HF spaces can inference in that ...
hey
its same as before
azure marshBOT
RVC Guides (How to Make AI Cover)
brittle wing
what should I rename these?
rare gobletBOT
Ayo? @brittle wing level 2 !!!
rigid light
featur_pit = self.compute_f0(inp_path, f0_method)
File "/mnt/personal/AI/Audio/Applio-RVC-Fork/lib/modules/train/extract/extract_f0_rmvpe.py", line 52, in compute_f0
f0 = self.model_rmvpe.infer_from_audio(x, thred=0.03)
File "/mnt/personal/AI/Audio/Applio-RVC-Fork/lib/modules/infer/rmvpe.py", line 660, in infer_from_audio
hidden = self.mel2hidden(mel)
File "/mnt/personal/AI/Audio/Applio-RVC-Fork/lib/modules/infer/rmvpe.py", line 641, in mel2hidden
hidden = self.model(mel)
File "/mnt/personal/AI/Audio/Applio-RVC-Fork/.venv/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1518, in _wrapped_call_impl
return self._call_impl(*args, **kwargs)
File "/mnt/personal/AI/Audio/Applio-RVC-Fork/.venv/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1527, in _call_impl
return forward_call(*args, **kwargs)
File "/mnt/personal/AI/Audio/Applio-RVC-Fork/lib/modules/infer/rmvpe.py", line 468, in forward
x = self.cnn(self.unet(mel)).transpose(1, 2).flatten(-2)
File "/mnt/personal/AI/Audio/Applio-RVC-Fork/.venv/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1518, in _wrapped_call_impl
return self._call_impl(*args, **kwargs)
File "/mnt/personal/AI/Audio/Applio-RVC-Fork/.venv/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1527, in _call_impl
return forward_call(*args, **kwargs)
File "/mnt/personal/AI/Audio/Applio-RVC-Fork/lib/modules/infer/rmvpe.py", line 426, in forward
x, concat_tensors = self.encoder(x)
File "/mnt/personal/AI/Audio/Applio-RVC-Fork/.venv/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1518, in _wrapped_call_impl
return self._call_impl(*args, **kwargs)
File "/mnt/personal/AI/Audio/Applio-RVC-Fork/.venv/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1527, in _call_impl
return forward_call(*args, **kwargs)
File "/mnt/personal/AI/Audio/Applio-RVC-Fork/lib/modules/infer/rmvpe.py", line 303, in forward
x = self.bn(x)
File "/mnt/personal/AI/Audio/Applio-RVC-Fork/.venv/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1518, in _wrapped_call_impl
return self._call_impl(*args, **kwargs)
File "/mnt/personal/AI/Audio/Applio-RVC-Fork/.venv/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1527, in _call_impl
return forward_call(*args, **kwargs)
File "/mnt/personal/AI/Audio/Applio-RVC-Fork/.venv/lib/python3.10/site-packages/torch/nn/modules/batchnorm.py", line 171, in forward
return F.batch_norm(
File "/mnt/personal/AI/Audio/Applio-RVC-Fork/.venv/lib/python3.10/site-packages/torch/nn/functional.py", line 2478, in batch_norm
return torch.batch_norm(
RuntimeError: Expected tensor for argument #1 'input' to have the same type as tensor for argument #2 'weight'; but type torch.cuda.FloatTensor does not equal torch.cuda.HalfTensor (while checking arguments for cudnn_batch_norm)
erm
stuck bane
``` File "/mnt/personal/AI/Audio/Applio-RVC-Fork/lib/modules/train/extract/extr...
same thing happens to me
shit is dumb asf
lime sorrel
why does it take too long to upload on huggingface?
stuck bane
``` File "/mnt/personal/AI/Audio/Applio-RVC-Fork/lib/modules/train/extract/extr...
been having this for like 3 days
and still no luck with reinstallling numerous times
brittle wing
basically in the easy rvc gui google colab thing when i do the second step which is to put the link to the model it doesnt work when its literally worked every time before
it says it couldnt access it even tho on google drive the model is set to everyone can access
violet heron
what do epochs mean
Training cycles, more/less does not mean better
brittle wing
it says it couldnt access it even tho on google drive the model is set to everyo...
it says this
finite galleon
it says it couldnt access it even tho on google drive the model is set to everyo...
Google drive links are broken rn
brittle wing
Google drive links are broken rn
wait wha why
finite galleon
You would need to switch to huggingface
finite galleon
wait wha why
Google added an antivirus so the script couldnt download it
brittle wing
You would need to switch to huggingface
any tuts for that or do i just make an account and upload it
brittle wing
Google added an antivirus so the script couldnt download it
:(
finite galleon
They just added it like about 3 days ago
finite galleon
any tuts for that or do i just make an account and upload it
-huggingface
azure marshBOT
-huggingface
Suggestions for @brittle wing
brittle wing
thank you
rigid light
been having this for like 3 days
I thought it was a problem with python3.11, but even downgrading it's still fuzucked
stuck bane
I thought it was a problem with python3.11, but even downgrading it's still fuzu...
delete miniconda
install windows installer
i think i got it to work
make sure u have vc red and build tools
and python 3.98
finite galleon
``` File "/mnt/personal/AI/Audio/Applio-RVC-Fork/lib/modules/train/extract/extr...
Does it work on other pitch extraction
stuck bane
Does it work on other pitch extraction
unfortunately its a runtime error
and probably some conda errors
it just doesnt install properly...
stuck bane
I thought it was a problem with python3.11, but even downgrading it's still fuzu...
did you install runtime
im installing it rn
and ill update if it fixed the issue
which i dont think it will...
lusty willow
I’m currently having issues when trying to download zip models via the Easy GUI.
Although I get the notification that the download is sucessful, when I go to the model inference, it’s not there. I refresh the voice list and index path, but still it doesn’t appear.
Furthermore, when I go to the GUI cell, this message pops up:
“Access denied with the following error:
Cannot retrieve the public link of the file. You may need to change the permission to 'Anyone with the link', or have had many accesses.”
However, when I go back to the zip folders, they are indeed set to “Anyone with the link”. So why am I getting this error?
Complicated to say the least, can anyone help me out here? 😕
rare gobletBOT
Ayo? @lusty willow level 2 !!!
rigid light
Does it work on other pitch extraction
it works on regular rmvpe, but it's way too slow for my use case
rigid light
did you install runtime
I don't use miniconda (wrong reply)
finite galleon
it works on regular rmvpe, but it's way too slow for my use case
Wait arent you using rmvpe on the error you sent
rigid light
Wait arent you using rmvpe on the error you sent
that's rmvpe_gpu
finite galleon
that's rmvpe_gpu
Ohh
finite galleon
that's rmvpe_gpu
Since when rmvpe gpu exists on applio
rigid light
Since when rmvpe gpu exists on applio
idk I used to use it until it broke
rigid light
best part is it doesn't matter what version I use
stuck bane
i guess the installer didn't install pytorch properly
rigid light
rigid light
i guess the installer didn't install pytorch properly
I even manually installed pytorch
stuck bane
I even manually installed pytorch
install pytorch with gpu support
pip install torch==1.10.0+cu113 torchvision==0.11.1+cu113 torchaudio==0.10.0+cu113 -f https://download.pytorch.org/whl/cu113/torch_stable.html
rigid light
why cu113
stuck bane
ill see if these work
rigid light
I already have
I've tried python3.11 with torch using cu117, cu118 and even the newest version
as well as python3.10, rinsed and repeat
rigid light
all of them
stuck bane
name them
rigid light
well I have python3-venv
instead of using python -m venv .venv to create the venv though, I have to use virtualenv --copies .venv
otherwise it throws symlink errors
stuck bane
i just have python 3.98, git, build tools, and then the other thing, and it installed miniconda thru the installer for me
and i installed runtime, but it still gives me a pytorch error, so if this works ill let u know
rigid light
I never have had miniconda
stuck bane
what system r u on
stuck bane
what even is debian
warm hawk
stuck bane
I could never figure out how to install it properly
so far so good
rigid light
bet he uses winrar
stuck bane
bet he uses winrar
dude my network keeps failing
maybe thats why my shi aint installing correctly
python doesnt have full access to my network
lime sorrel
hello, i got this error
the thing is i trained using the RVC v2 colab
so this shouldn't even happen
silver bobcat
im getting warn cpu in hina ai gen cover can someone help me fix?
merry cedar
forest sluice
im getting warn cpu in hina ai gen cover can someone help me fix?
Edit > Notebook Settings
See if "T4 GPU" is selected
If not, select it
lime sorrel
is there anything i did wrong?
rigid light
Look at CUDA Version
CUDA Version: 12.0
rigid light
I'm runnin an RTX 3090
forest sluice
And are you running on a VM?
rigid light
Yes
forest sluice
I looked at your home dir, it says "personalvm"
stuck bane
im on rtx 3070
winter holly
Mine wont do the conversion at all, how do I run properly? I am on a 3060 with the most recent version on hugging.
rigid light
Proxmox
rare gobletBOT
Ayo? @winter holly level 1 !!!
rigid light
I have a server computer I put together running multiple VMs, this one primarily for AI
stuck bane
return F.batch_norm(
File "C:\Users\chrus\Desktop\Applio-RVC-Fork\env\lib\site-packages\torch\nn\functional.py", line 2478, in batch_norm
return torch.batch_norm(
RuntimeError: Expected tensor for argument #1 'input' to have the same type as tensor for argument #2 'weight'; but type torch.cuda.FloatTensor does not equal torch.cuda.HalfTensor (while checking arguments for cudnn_batch_norm)
rigid light
I've been moving my various AI things over to it, and so far everything has worked except for RVC
stuck bane
do yk how to fix @forest sluice
rigid light
I've gotten Stable Diffusion and UVR5 working no problem, but this one makes me want to 
forest sluice
return F.batch_norm(
File "C:\Users\chrus\Desktop\Applio-RVC-Fork\env\lib\site...
That's library error I'm afraid
stuck bane
That's library error I'm afraid
what do i need
rigid light
That's library error I'm afraid
well
stuck bane
we get the same error
winter holly
@forest sluice Mine wont do the conversion at all, how do I run properly? I am on a 3060 with the most recent version on hugging. How do I fix this?
rigid light
I GOT that error in python 3.11 and 3.10 in Applio
stuck bane
mines 3.98
rigid light
on the og RVC repo, I get this crappy socket error
stuck bane
i tried like 7 repo's
they all do the same thing
either a runtime, torch, or a tensorboard error
there's no in between
nothing else
stuck bane
That's library error I'm afraid
what do u recommend
forest sluice
well
https://developer.nvidia.com/cuda-12-1-0-download-archive?target_os=Linux&target_arch=x86_64&Distribution=Debian&target_version=11 Install cuda 12.1, reboot, then install pytorch, reboot again?
rigid light
something's not being passed correctly to the rmvpe .py
rigid light
https://developer.nvidia.com/cuda-12-1-0-download-archive?target_os=Linux&target...
It's already installed, also I can't install that as it's targeting Debian 11
stuck bane
https://developer.nvidia.com/cuda-12-1-0-download-archive?target_os=Linux&target...
same for me?
do i need that?
rigid light
On Debian 12
stuck bane
do i need that except for windows
@forest sluicei got rtx 3070, windows 11, which cuda toolkit do i need, and which pytorch version
forest sluice
It's already installed, also I can't install that as it's targeting Debian 11
That's gonna be painful. Have you launched uvr5 again?
forest sluice
<@895958242105122826>i got rtx 3070, windows 11, which cuda toolkit do i need, a...
nvidia-smi
forest sluice
https://discuss.pytorch.org/t/question-i-have-a-question-about-installing-pytorch/191829/2 Oh, pytorch just needs an NVIDIA GPU
stuck bane
https://discuss.pytorch.org/t/question-i-have-a-question-about-installing-pytorc...
yet i have installed the latest pytorch and still not success
rare gobletBOT
Ayo? @stuck bane level 9 !!!
rigid light
That's gonna be painful. Have you launched uvr5 again?
yes, I used it to prep my dataset
stuck bane
rigid light
SD still works
forest sluice
Maybe downgrade to Cuda 12.1 then
rigid light
I'm already using cu118 for UVR5, SD uses cu121
forest sluice
Maybe downgrade to Cuda 12.1 then
I can't provide much help, as I don't have an NVIDIA GPU locally
rigid light
also wdym downgrade
stuck bane
kk ill try that, for a sec i thought you meant on pytorch
oh that's the latest it supports
HAHA
found my issue i think
thanks
rigid light
File "/usr/lib/python3.11/multiprocessing/queues.py", line 244, in _feed
obj = _ForkingPickler.dumps(obj)
^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/lib/python3.11/multiprocessing/reduction.py", line 51, in dumps
cls(buf, protocol).dump(obj)
File "/mnt/personal/AI/Audio/Retrieval-based-Voice-Conversion-WebUI/.venv/lib/python3.11/site-packages/torch/multiprocessing/reductions.py", line 429, in reduce_storage
df = multiprocessing.reduction.DupFd(fd)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/lib/python3.11/multiprocessing/reduction.py", line 198, in DupFd
return resource_sharer.DupFd(fd)
^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/lib/python3.11/multiprocessing/resource_sharer.py", line 53, in __init__
self._id = _resource_sharer.register(send, close)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/lib/python3.11/multiprocessing/resource_sharer.py", line 76, in register
self._start()
File "/usr/lib/python3.11/multiprocessing/resource_sharer.py", line 126, in _start
self._listener = Listener(authkey=process.current_process().authkey)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/lib/python3.11/multiprocessing/connection.py", line 447, in __init__
self._listener = SocketListener(address, family, backlog)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/lib/python3.11/multiprocessing/connection.py", line 590, in __init__
self._socket.bind(address)
PermissionError: [Errno 1] Operation not permitted
welp
forest sluice
first, use python 3.10. 3.11 just don't work
rigid light
first, use python 3.10. 3.11 just don't work
Well shit, I'll have to completely resetup my 3.10 install
stuck bane
Well shit, I'll have to completely resetup my 3.10 install
Create a separate environment if you can
stuck bane
first, use python 3.10. 3.11 just don't work
Resetting my computer
It ain’t even letting me extract the toolkit to download
Lmaooo
Fucking applio man, gonna do all this shit just for them to release a whole new client next week
rigid light
Create a separate environment if you can
I don't have conda
stuck bane
I don't have conda
If you have patience wait for beta testing on Monday
rigid light
I don't think they're gonna fix the socket issue
zinc anchor
first, use python 3.10. 3.11 just don't work
Why is that?
stuck bane
Yeah it’s a whole new client
forest sluice
```Traceback (most recent call last):
File "/usr/lib/python3.11/multiprocessin...
Are you still running A1111?
stuck bane
https://stackoverflow.com/questions/11145463/python-socket-error-operation-not-p...
Chat gpt gives a diff answer
forest sluice
User actually ran proxmox, and I don't know whatever the hell does proxmox do in Client-Socket intervention
Might as well running sudo as a last ditch effort
rigid light
Are you still running A1111?
yes
rigid light
Chat gpt gives a diff answer
do not trust + verify chatgpt answers
carmine osprey
-rvc
stuck bane
@forest sluice quick question, do you even need to install nvidia runtime for applio in order to train with gpu
what will it even do
brittle wing
Thank you
rigid light
@forest sluice
File "/home/personalvm/.pyenv/versions/3.10.13/lib/python3.10/multiprocessing/queues.py", line 244, in _feed
obj = _ForkingPickler.dumps(obj)
File "/home/personalvm/.pyenv/versions/3.10.13/lib/python3.10/multiprocessing/reduction.py", line 51, in dumps
cls(buf, protocol).dump(obj)
File "/home/personalvm/.pyenv/versions/3.10.13/envs/.venv/lib/python3.10/site-packages/torch/multiprocessing/reductions.py", line 429, in reduce_storage
df = multiprocessing.reduction.DupFd(fd)
File "/home/personalvm/.pyenv/versions/3.10.13/lib/python3.10/multiprocessing/reduction.py", line 198, in DupFd
return resource_sharer.DupFd(fd)
File "/home/personalvm/.pyenv/versions/3.10.13/lib/python3.10/multiprocessing/resource_sharer.py", line 53, in __init__
self._id = _resource_sharer.register(send, close)
File "/home/personalvm/.pyenv/versions/3.10.13/lib/python3.10/multiprocessing/resource_sharer.py", line 76, in register
self._start()
File "/home/personalvm/.pyenv/versions/3.10.13/lib/python3.10/multiprocessing/resource_sharer.py", line 126, in _start
self._listener = Listener(authkey=process.current_process().authkey)
File "/home/personalvm/.pyenv/versions/3.10.13/lib/python3.10/multiprocessing/connection.py", line 448, in __init__
self._listener = SocketListener(address, family, backlog)
File "/home/personalvm/.pyenv/versions/3.10.13/lib/python3.10/multiprocessing/connection.py", line 591, in __init__
self._socket.bind(address)
PermissionError: [Errno 1] Operation not permitted
Same result
stiff goblet
Can i use new pretrained models for 32khz training?
stuck bane
Can i use new pretrained models for 32khz training?
Yeah but what is a khz
forest sluice
sample rate
rigid light
IT BEGINS
I had to completely butcher my virtual environment and .bashrc BUT I DID IT
reef pasture
2 things:
But how do you create a voice like you do?
How do you speak with those voices or take a vowel and put those voices?
scarlet relic
Bringing this up again cuz i rlly need help
sand iris
I had to completely butcher my virtual environment and .bashrc BUT I DID IT
What happened there ?
proper shale
Does CMD say anything
proper shale
2 things:
But how do you create a voice like you do?
How do you speak with th...
-rvc
proper shale
These guides should explain what you need to know
forest sluice
What happened there ?
#✨│ai-help message Basically a bunch of stupid shenanigans
sand iris
Ah nevermind. I thought this was training under radeon cards
Still can't figure out why it becomes so slow for no reason sometimes
elder cape
hello
forest sluice
Still can't figure out why it becomes so slow for no reason sometimes
Oh, and btw, tkinter is baked into a typical Python install, so tk package only contains C library & GUI
elder cape
I am a korean.
I'm currently training voices using AI and making songs with it, but an error suddenly occurred in Colab. (Error PTH file not found.)
Can you tell me why?
rare gobletBOT
Ayo? @elder cape level 1 !!!
rigid light
What happened there ?
I have to enter root to prevent the OS screaming at me about opening ports, but in order to access the python container for 3.10 (pyenv), pyenv usually can only run in non-root situations
so I had to force run pyenv under root, then create the virtual environment for RVC under root, and then enter it as root
While also getting pyenv to be detected while I'm in root (that's where the bashrc fuckery comes in)
elder cape
Sorry, but can you lower your English to kindergarten level?
I am a korean. I do not speak English well. Sorry.
I was making a song using Colab.
Starting yesterday, education through Colab suddenly stopped working.
I entered and proceeded normally, but the same error appears.
Google Drive - Share - No link restrictions - Step 2. An error continues to appear in Download The Model. PTH file not found. saying..
Is there any way to solve this?
worn glacier
How to make RVC TTS? I can make only songs with Colab, not the TTS, how do i do that?
shy dagger
Hello
forest sluice
Sorry, but can you lower your English to kindergarten level?
I am a korean. I do...
Give me model download link
azure marshBOT
RVC Guides (How to Make AI Cover)
scarlet relic
Does CMD say anything
just a lot of this stuff without stopping, now i can somewhat hear something but its like messed up noises, not words and its almost never
forest sluice
https://drive.google.com/file/d/1mZJkW7Z2vuAMS5E5wrK1ePtpBeMoJ8Di/view?usp=shari...
You made your own model? You forgot to train index btw.
elder cape
Give me model download link
I made it myself.
What is index btw?
The PTH file uploaded by someone else also did not work properly. If there is a currently working file, can you please provide a link?
scarlet relic
Does CMD say anything
I either missed it last time or it didnt happen, but when i closed it now i got this error
proper shale
I either missed it last time or it didnt happen, but when i closed it now i got ...
That should be normal technically
But idk, I suggest using W-Okada for realtime
runic stump
I’m currently having issues when trying to download zip models via the Easy GUI....
Same exact thing happened to me. Came on here to see if anyone else had the issue. Regardless of whether it's one of my models that's worked in the past or someone else's, everything is giving me the same error.
runic stump
Do any of these work with already completed RVC V2 zip models? I went through them earlier and couldn't find what I was looking for on the colab side. I tried the huggingface one and couldn't get that one to load zip models either (unless I just don't know what I'm doing, it's very late where I am, maybe i just need sleep first)
proper shale
Do any of these work with already completed RVC V2 zip models? I went through th...
Well, there aren't any mainstream inferencing Colabs.
The Huggingface spaces are good, however you need to upload your models to Huggingface because G Drive models got broken
They're way slower in inferencing speed though, because it's CPU only
runic stump
Alright, I'll look into that more when I wake up. Thanks for your help! Do you know if there are plans to fix Drive models soon?
scarlet relic
But idk, I suggest using W-Okada for realtime
I do, i just wanted to try mangio since i was told it might perform equal or better in some cases
proper shale
Alright, I'll look into that more when I wake up. Thanks for your help! Do you k...
Well, maybe it was some Google update, I believe the devs are working for a fix but I'm unsure.
And you're welcome :)
proper shale
I do, i just wanted to try mangio since i was told it might perform equal or bet...
...but it doesn't
- W-Okada has more options
runic stump
Well, maybe it was some Google update, I believe the devs are working for a fix ...
Thanks so much, have a great day :)
rare gobletBOT
Ayo? @runic stump level 1 !!!
proper shale
Thanks so much, have a great day \:)
You too!!!
scarlet relic
Np, sorry if I couldn't help enough
nah dw i appreciate it!
proper shale
did u open start https? hmm
rare gobletBOT
Ayo? @brittle wing level 1 !!!
proper shale
u should use start_http yeah
hmmm i honestly have no clue
check this though, might help: https://rentry.co/W-Okada-FAQ
gusty kestrelBOT
I have found 2 results for the search skaven...
Requested by dobrynya
I have found 1 results for the search spacemarine...
Requested by dobrynya
tired elk
Oops, wrong channel
proper shale
It's alright
frosty night
why does latest rvc realtime gui freeze and crash when i click "start audio conversion"
minor juniper
Who can tell me, when playing a voice in Voice changer it makes a Trale sound
How to remove this sound during a conversation
finite flicker
Anyone else unable to select downloaded models in RVC v2 huggingface..? Even after refreshing
proper shale
Anyone else unable to select downloaded models in RVC v2 huggingface..? Even aft...
Is the model's download link from Google drive?
If so, I suggest downloading the zip and reuploading as a model on Huggingface
finite flicker
Is the model's download link from Google drive?
That's right
finite flicker
If so, I suggest downloading the zip and reuploading as a model on Huggingface
Thank you, I will try it ❤️
proper shale
-huggingface
azure marshBOT
-huggingface
finite flicker
wow, thank you so much!!!
stray iris
training voice model in colab, then download .index and .py files to the properly directories. Local web-ui can see my model, but feed error
Traceback (most recent call last):
File "D:\RVC_and_others\RVC-beta0717\runtime\lib\site-packages\gradio\routes.py", line 321, in run_predict
output = await app.blocks.process_api(
File "D:\RVC_and_others\RVC-beta0717\runtime\lib\site-packages\gradio\blocks.py", line 1007, in process_api
data = self.postprocess_data(fn_index, result["prediction"], state)
File "D:\RVC_and_others\RVC-beta0717\runtime\lib\site-packages\gradio\blocks.py", line 953, in postprocess_data
prediction_value = block.postprocess(prediction_value)
File "D:\RVC_and_others\RVC-beta0717\runtime\lib\site-packages\gradio\components.py", line 2076, in postprocess
processing_utils.audio_to_file(sample_rate, data, file.name)
File "D:\RVC_and_others\RVC-beta0717\runtime\lib\site-packages\gradio\processing_utils.py", line 206, in audio_to_file
data = convert_to_16_bit_wav(data)
File "D:\RVC_and_others\RVC-beta0717\runtime\lib\site-packages\gradio\processing_utils.py", line 219, in convert_to_16_bit_wav
if data.dtype in [np.float64, np.float32, np.float16]:
AttributeError: 'NoneType' object has no attribute 'dtype'
what`s wrong?
Do i need download some more files from drive?
okaaay, magic reboot fix that lmao
finite galleon
> Traceback (most recent call last):
> File "D:\RVC_and_others\RVC-beta0717\ru...
Is your audio file path correct
Never mind
finite galleon
okaaay, magic reboot fix that lmao
I am so fucking blind at night that i didnt see this
stray iris
I am so fucking blind at night that i didnt see this
Thanks anyway
rare gobletBOT
Ayo? @stray iris level 4 !!!
harsh island
im having trouble adding a model in the voice changer can someone h elp me
File "restapi\MMVC_Rest_Fileuploader.py", line 92, in post_load_model
File "voice_changer\VoiceChangerManager.py", line 161, in loadModel
File "voice_changer\RVC\RVCModelSlotGenerator.py", line 37, in loadModel
File "voice_changer\RVC\RVCModelSlotGenerator.py", line 43, in _setInfoByPytorch
KeyError: 'config'
paramDict {'voiceChangerType': 'RVC', 'slot': 20, 'isSampleMode': False, 'sampleId': None, 'files': [{'name': 'G_Snowie_RuPretrain_EnP.pth', 'kind': 'rvcModel', 'dir': ''}], 'params': {}}
RVC:: slotInfo.modelFile G_Snowie_RuPretrain_EnP.pth
[Voice Changer] post_load_model ex: 'config'
Traceback (most recent call last):
File "restapi\MMVC_Rest_Fileuploader.py", line 92, in post_load_model
File "voice_changer\VoiceChangerManager.py", line 161, in loadModel
File "voice_changer\RVC\RVCModelSlotGenerator.py", line 37, in loadModel
File "voice_changer\RVC\RVCModelSlotGenerator.py", line 43, in _setInfoByPytorch
KeyError: 'config'```You probably don't want to upload that model 😓
From what I can guess from the name, it's the pretrain which is meant to create new models but not change voice
dusk brook
harsh island
From what I can guess from the name, it's the pretrain which is meant to create ...
OHHH okay thank you!
grave plaza
hey i had a quick question for yall, is there any benefits of using ngrok over gradio?
rare gobletBOT
Ayo? @grave plaza level 1 !!!
nocturne mural
question , who has had problems with google drive links?
rare gobletBOT
Ayo? @nocturne mural level 5 !!!
molten pecan
question , who has had problems with google drive links?
Creo que @brittle wing
nocturne mural
Creo que <@456226577798135808>
eso lo se
but this chat is english
and I was just asking because I fixed the problem
molten pecan
and I was just asking because I fixed the problem
Doing god's work 🙏
Good job Vidal
nocturne mural
because google now puts to the files the error Google Drive cannot scan this file for viruses.
molten pecan
because google now puts to the files the error Google Drive cannot scan this fil...
Is it even good at detecting viruses?
stray swift
I'm I overtraining ? 400 epochs..
nocturne mural
Is it even good at detecting viruses?
no idea, but that only happened if the file was very heavy, but it seems that this limit has been lowered.
molten pecan
no idea, but that only happened if the file was very heavy, but it seems that th...
Interesting
grave plaza
hey i had a quick question for yall, is there any benefits of using ngrok over g...
:( is there?
stray swift
I don't think so
I don't think so either, but damn 400 epochs and still going ? Weird to me
ionic roost
Damn your chart is hella smooth
stray swift
Damn your chart is hella smooth
Thanks, spends some weeks cleaning the dataset 
rare gobletBOT
Ayo? @ionic roost level 4 !!!
stray swift
nocturne mural
en general todos los q utilizan la dependencia gdown
solo porque la dependencia no estaba preparado ante eso
stray swift
Looks like it's smoothed out right?
I will wait 30 min just to sure
half cove
question , who has had problems with google drive links?
alot of people have
nocturne mural
alot of people have
I doubt if others know how to fix it in their colabs.
half cove
nocturne mural
but for now I already fixed it in the applio
but if you want I can fix the others
half cove
half cove
yes or not?
yes 
nocturne mural
Which one do I start with or is it more commonly used here?
half cove
Which one do I start with or is it more commonly used here?
ila rvc
ig
and uhh
eddys colab
noui
nocturne mural
half cove
proven hill
GOTH GIRLFRIEND VOICE MODEL 
Finally the Big T- Goth Mommy Girlfriend voice model is here! 
Grab your copy here now: https://ko-fi.com/s/115b3d9742
Or contact me in DM for a 20% discount! 
azure marshBOT
RVC Guides (How to Make AI Cover)
torn wasp
-colab
spark adder
the RVC Crepe v2 colab gives me an "access denied" message when trying to load a model, despite the model being public and working beforehand, anyone else had this? It does the same when trying to use a reupload
half cove
the RVC Crepe v2 colab gives me an "access denied" message when trying to load a...
with google drive?
spark adder
Yes
half cove
spark adder
Oh my bad, thank u !
molten pecan
@torn wasp que querías saber?
torn wasp
<@839676250725154876> que querías saber?
Ya te comento jaja
dusky fiber
brittle wing
any RVC which works on Mac M1 yet?
sand iris
any RVC which works on Mac M1 yet?
they should be working already
model training doesn't though from what I've been told
mint tendon
what's the difference between this one and w-okada's?
half cove
what's the difference between this one and w-okada's?
this is for converting audio okada is for realtime
think of it like
this for audio files
proper shale
this is for converting audio okada is for realtime
both use the RVC AI tho
half cove
both use the RVC AI tho
yea but i assume
they mean
also the go realtime on things that arnt okada is goofy
proper shale
also the go realtime on things that arnt okada is goofy
yeye
molten pecan
@torn wasp La página regresa a sus valores predeterminados cuando reinicias el espacio.
- Ve a Settings arriba a la derecha
- Baja hasta encontrar
Restart this Space - Toca restart a la derecha. Volverá a cargar de cero
mint tendon
this is for converting audio okada is for realtime
ooh okay, thank you
high nebula
alguien con el modelo de randy? de jowell y randy
spare shale
alguien con el modelo de randy? de jowell y randy
No lo tengo, pero intenta encontrarlo en #1175430844685484042
O pedir una voz
agile lynx
Whats epochs meaning?
acoustic light
How do you make backing vocals change to the voice models voice, I put the instrumental + backing vocals into rvc and it isnt changing
acoustic light
Whats epochs meaning?
its like pitch
agile lynx
its like pitch
Hm okay
violet heron
its like pitch
No
@agile lynx an epoch is how much the AI looks over the dataset
So like if a model was 300 epochs the AI looked over it 300 times
More/less does not mean better
agile lynx
More/less does not mean better
Like when i set 1000 its bad and 1 too?
violet heron
Like when i set 1000 its bad and 1 too?
Really just depends on the dataset
Many people train 250-500
fathom fog
I was wondering if anybody else was having issues downloading their voice models to spaces for voice conversion like the Illaria RVC tool on hugging space or any others? None of them are downloading my voice models no matter how many times I might download and refresh the model list.
violet heron
I was wondering if anybody else was having issues downloading their voice models...
What’s the model link
fathom fog
What’s the model link
https://huggingface.co/spaces/Clebersla/RVC_V2_Huggingface_Version
https://huggingface.co/spaces/TheStinger/Ilaria_RVC
Neither or these spaces and a few others work for me anymore, just using direct google drink links to the pth zip.
violet heron
https://huggingface.co/spaces/Clebersla/RVC_V2_Huggingface_Version
https://hugg...
That’s why
Drive link
stray iris
Are there any guides to improve the output audio? Cause I trained voice model today, everything is ok, but when I let it revoice song, it sound absolutle bad. Yeah, we can hear some words in song and voice model really works, buuuuut sound..............
violet heron
Drive link
Upload to mega, HuggingFace, or Dropbox
fathom fog
That’s why
Ah thanks, the drive links worked until today for me specifically, any reason for that? I'm going to change but just curious yknow.
rare gobletBOT
Ayo? @fathom fog level 1 !!!
violet heron
Ah thanks, the drive links worked until today for me specifically, any reason fo...
Google broke it
young jasper
Hope it's the right channel and doesn't sound weird, but has anyone had the issue when using rvcs in a voice changer that most of the time it seems to "add" a slight British accent? I'm not British and not using voice models of British people...
errant helm
rare gobletBOT
Ayo? @spare shale level 17 !!!
molten pecan
@hexed panther we don't recommend RVC GUI
It's really outdated
Use Mainline or Mangio
-rvc
molten pecan
Check the links collection post on #1159513888199540817
small umbraBOT
https://discord.com/channels/1159260121998827560/1175430844685484042
/search https://discord.com/channels/1159260121998827560/1163592055830880266
or take a look at weights.gg
do not use `/search` outside of #search-models
copper acorn
How do you properly merge models in rvc? Everytime I've tried I got a result that sounds extremely broken
stray iris
Use Mainline or Mangio
Is there collab(or something else) with mangio?
violet heron
Is there collab(or something else) with mangio?
Used to
Google of course had to butcher it
stray iris
Damn
molten pecan
Google of course had to butcher it
Uhh aicovergen is built over mangio
timber apex
GOD I NEED HELP
molten pecan
Or at least that's what colab says
If you look over the internal files it says "mangio-rvc-fork"
Plus it has mangio-crepe too
violet heron
Uhh aicovergen is built over mangio
All of the colabs are
timber apex
I try to put my dataset name in rvc disconnected ant it gave me an error that it didn't find it
violet heron
I try to put my dataset name in rvc disconnected ant it gave me an error that it...
Where is the dataset located
molten pecan
I try to put my dataset name in rvc disconnected ant it gave me an error that it...
Where did you put the dataset on?
timber apex
im gonna get the dataset 1 sec...
half cove
Uhh aicovergen is built over mangio
isnt it based of mainline?
molten pecan
isnt it based of mainline?
Apparently it isn't. Check the folders
proper shale
I try to put my dataset name in rvc disconnected ant it gave me an error that it...
What's the name of your model
timber apex
whats a experiment name?
vapid wren
Hey, can someone help? I'm trying to do the training.
proper shale
Hey, can someone help? I'm trying to do the training.
Apparently you didn't feature extract
proper shale
whats a experiment name?
The name of the model
timber apex
shirtsandpants.zip
proper shale
The name of the model
the model
vapid wren
Apparently you didn't feature extract
Meaning?
timber apex
thats the model
proper shale
thats the model
That's the dataset name, I'm asking about the model name in case it has spaces
vapid wren
Oh
timber apex
That's the dataset name, I'm asking about the model name in case it has spaces
SHIRTSANDPANTS IS THE MODEL bruh
copper acorn
Is merging models easy or is my rvc just broken?
glass garden
Can someone help me out with using a voice for a project
I’m trying to use BMO from adventure time but I don’t know how to
vapid wren
How long does training take?
proper shale
SHIRTSANDPANTS IS THE MODEL bruh
Did all the Dependencies steps run correctly
timber apex
That's the dataset name, I'm asking about the model name in case it has spaces
is it like the folder named "rvcDisconnected" or the zip inside the folder
proper shale
How long does training take?
This depends on the batch_size and dataset size.
You should limit your training sessions to 1/2hrs on colab
glass garden
How do I set this up
rare gobletBOT
Ayo? @glass garden level 1 !!!
vapid wren
First time using this
Same
proper shale
is it like the folder named "rvcDisconnected" or the zip inside the folder
I was asking about what you inserted in "experiment_name" under the Set Training Variables step.
proper shale
How do I set this up
What program specifically
timber apex
"shirtsandpants"?
vapid wren
Oh, what's this now? Is it a bad thing?
timber apex
yeah i did that
glass garden
Tbh I don’t know what to do I mainly want to train an ai to sound like BMO from adventure time
proper shale
What program specifically
There's the voice changer and the audio inferencing/training models
proper shale
Tbh I don’t know what to do I mainly want to train an ai to sound like BMO from ...
-rvc
half cove
Oh, what's this now? Is it a bad thing?
thats tensorboard
proper shale
check Create a model with RVC disconnected
glass garden
Thank you
proper shale
Oh, what's this now? Is it a bad thing?
Nah not bad, click on refresh once it reaches like 25 epochs
proper shale
i did ||shirtsandpants||
Hmm this is weird
vapid wren
Nah not bad, click on refresh once it reaches like 25 epochs
Alr.
timber apex
why
proper shale
Are you sure that both the Colab and the Drive are using the same account
vapid wren
Wait what
timber apex
Hmm this is weird
Why?
vapid wren
It has to do every single Epoch? Damn, my patience..
Yeah no, Ima do this tomorrow
Literally 11 pm for me rn
I suppose this will take hours
proper shale
I suppose this will take hours
If you wanna train it fully, it's gonna take hours in total
You can train a lil bit every day though
timber apex
Because it's not a common error
IN THIS STEP BRUH
proper shale
IN THIS STEP BRUH
I'm aware... I'm just saying that this ain't a common thing
timber apex
it says cannot find the dataset in rvcdisconnected folder
proper shale
IN THIS STEP BRUH
...Now, you insert the name of your zip file, and run that step
timber apex
...Now, you insert the name of your zip file, and run that step
it says cannot find the dataset in rvcdisconnected folder
vapid wren
How can I be able to not lose my training progress?
timber apex
did you write .zip at the end
yeah maybe
proper shale
How can I be able to not lose my training progress?
You can stop after a saving epoch
half cove
yeah maybe
maybe?
timber apex
yes
timber apex
what why
timber apex
whats uhh?
vapid wren
You can stop after a saving epoch
After a saving epoch?
timber apex
yes
molten pecan
Did they contain spaces or special characters?
timber apex
as .zip
proper shale
After a saving epoch?
Yeah, you set it to every 25 epochs as far as I can tell
timber apex
yeah
vapid wren
Yeah, you set it to every 25 epochs as far as I can tell
Wait what, how though?
half cove
timber apex
THE AUDIOS HAVE TO BE ZIP INSTEAD OF WAV?!
proper shale
Wait what, how though?
Save frequency option
half cove
THE AUDIOS HAVE TO BE ZIP INSTEAD OF WAV?!
if its in a zipped folder
then it will automatically be zipped
half cove
timber apex
what
molten pecan
Did they contain spaces or special characters?
Do they?
vapid wren
Save frequency option
Am I weird for not knowing where that option is?
jaunty shale
It is there, but still showing up this error. Any ideas?
rare gobletBOT
Ayo? @jaunty shale level 5 !!!
timber apex
@half cove ? You there
half cove
It *is* there, but still showing up this error. Any ideas?
dont use a space
half cove
<@936673139419664414> ? You there
its installing the uhh
timber apex
what?
jaunty shale
dont use a space
ooh ok
half cove
Dependencies
uneven nacelle
So I was streaming with the rvc. And I had rtx turned on. And at some point when my game stuttered the program ceased to work. I guess my gpu just couldnt handle all of that work. But is there a way to get around this? Like how can I be sure that the program doesnt break?
timber apex
Dependencies
ok
proper shale
Am I weird for not knowing where that option is?
It's in the Training step
half cove
Dependencies haunt my dreams after that ov2super 32k shit
proper shale
So I was streaming with the rvc. And I had rtx turned on. And at some point when...
Don't use RTX with it, basically.
Make sure you're saving some power for the voice changer too
vapid wren
It's in the Training step
Oh, it's set to 25. So when I casually close my browser and open colab it's saved?
proper shale
It *is* there, but still showing up this error. Any ideas?
Model name has spaces. Remove the space
uneven nacelle
Don't use RTX with it, basically.
Make sure you're saving some power for the vo...
Alright but I limited the fps.
half cove
i already told em
timber apex
Dependencies haunt my dreams after that ov2super 32k shit
reply me when you want to
proper shale
Oh, it's set to 25. So when I casually close my browser and open colab it's save...
No, you need to stop training after a saving epoch, and export the model using the export model step
proper shale
You're welcome :)
half cove
reply me when you want to
it works for me 😭
half cove
that shouldnt matter
timber apex
ok
proper shale
i used edge and it gave me a error
This most likely isn't the reason for the error
timber apex
that shouldnt matter
after you are done, give me the zip with the index and pth
proper shale
I'm thinking it could be maybe just G Drive being broken or something, or the Load dataset step being broken 😭
half cove
timber apex
<:drakepain:1159570745068888174>
oh plus private chat me so you can give me the zip
half cove
timber apex
<:drakepain:1159570745068888174>
not public chat the zip so everyone can use it ok?
timber apex
<:boohooh:1176674698629750975>
what happen
timber apex
show me the screenshot bruh💀
half cove
show me the screenshot bruh💀
of?
vapid wren
No, you need to stop training after a saving epoch, and export the model using t...
Exported, when I want to continue, I do what? Just go back to the site or? Sorry I don't really know how this works
proper shale
Exported, when I want to continue, I do what? Just go back to the site or? Sorry...
Export once you step the training step
stop*
woops lol
half cove
the dataset that you are struggling
it has multiple speakers
proper shale
H wait im dumb
half cove
and uhh questionable quality
timber apex
it has multiple speakers
i put it at mono not stereo
timber apex
mono audio
proper shale
Exported, when I want to continue, I do what? Just go back to the site or? Sorry...
Alright so once you want to continue, you run all the steps until Set Training Variables normally, insert the same settings there.
Then you skip to the steps that are marked like this: (for resuming)
Then just run the training step with the same batch size as before and it should be retraining!
Or well, continuing to train
People use the term retraining for that lol
timber apex
then just choose person that you like and then get the split the vocals for the person and then i will have a good ai model
sounds good @half cove ?
half cove
😭
vapid wren
Alright so once you want to continue, you run all the steps until Set Training V...
I think I get it
proper shale
I think I get it<:nails:1159569314848972891>
If you have any issues once continuing, just lmk, ping me tmrw
;)
vapid wren
Alr thx
timber apex
WHAT IS IT THIS TIME AGAIN bruh @half cove
half cove
your dataset wont make a good model
timber apex
what will it sound like??
half cove
like shit
timber apex
show me a audio what will it sound like
and increase the quality so it can be better
vapid wren
If you have any issues once continuing, just lmk, ping me tmrw
I just want to double check if I can really close my browser now 
If an unzipped file shows up with the same name(that I didn't put in there) , does it mean that I've exported it?
proper shale
Check if the D/G files are there
lean owl
🔥
proper shale
Yeah that's it, ya can close the thing
timber apex
@half cove ?
half cove
😭
timber apex
What
timber apex
could you turn it into a good model
rare gobletBOT
Ayo? @timber apex level 7 !!!
half cove
no
stray iris
just use this -><:FL:1159653069743726664>and increase the quality so it can be b...
The biggest problem isn't a quality. RVC do a mix of this voices, cause it can't split it. Just mix. And it will sounds horrible
timber apex
how do i make it into a good model??
stray iris
how do i make it into a good model??
Split voices by yourself in other program
timber apex
Split voices by yourself in other program
wait a minute, HOW DO YOU KNOW ALL OF THIS IF YOU ARE NOT VERIFIED
stray iris
Cause it's basic?
proper shale
💀
lunar gale
when i talk with a voice model it is alway saing it again like in a loop but getting quiter each time how do i fix this?
proper shale
when i talk with a voice model it is alway saing it again like in a loop but get...
Activate Sup2, increase S thresh
What you described was echoing
timber apex
im now doing 1 speaker now...
azure marshBOT
RVC Guides (How to Make AI Cover)
brittle wing
tensorLaunch.py", line 2, in <module>
from tensorboard import program
ModuleNotFoundError: No module named 'tensorboard'
any help
fluid topaz
tensorLaunch.py", line 2, in <module>
from tensorboard import program
Module...
need to install tensorboard first
@echo off
setlocal
REM Check if tensorboard_venv exists in the root directory
if not exist "%~dp0\tensorboard_venv\" (
REM Print the step to the terminal
echo Installing virtualenv...
REM Install virtualenv
%~dp0\runtime\python -m pip install virtualenv
REM Print the step to the terminal
echo Creating a virtual environment named tensorboard_venv...
REM Create a virtual environment named tensorboard_venv in the root directory
%~dp0\runtime\python -m virtualenv %~dp0\tensorboard_venv
REM Print the step to the terminal
echo Activating the virtual environment...
REM Activate the virtual environment
call %~dp0\tensorboard_venv\Scripts\activate
REM Print the step to the terminal
echo Installing TensorBoard into the virtual environment...
REM Install TensorBoard into the virtual environment
pip install tensorboard
REM Downgrade problematic packages
echo Downgrading packages for troubleshooting...
pip install markdown==3.0
pip install tensorboard==2.1.0
pip install protobuf==3.11.0
pip install numpy==1.19.5
) else (
REM Print the step to the terminal
echo tensorboard_venv already exists, skipping creation and activation...
REM Activate the existing virtual environment
call %~dp0\tensorboard_venv\Scripts\activate
)
REM Print the step to the terminal
echo Launching TensorBoard...
REM Launch TensorBoard
tensorboard --logdir="%~dp0\logs"
REM Print the step to the terminal
echo Keeping the command prompt open...
pause
endlocal
i instlled ready
brittle wing
need to install tensorboard first
```
@echo off
setlocal
REM Check if tensorbo...
bat filer ?
fluid topaz
bat filer ?
yea its for .bat
brittle wing
yea its for .bat
ok
fluid topaz
i instlled ready
i think you installed it globally, while the rvc need it in virtual environment
brittle wing
i think you installed it globally, while the rvc need it in virtual environment
so the coment will install tensorboard on virtual environment right
fluid topaz
right, run inside the rvc folder
tight ibex
idk if this question suites this channel but is vcclient legit?
sand iris
yes it is
tight ibex
thank you
rare gobletBOT
Ayo? @tight ibex level 1 !!!
brittle wing
@fluid topaz
jaunty shale
I do not know if this is the appropiate chat for this topic, but I know instrumental is not allowed for audio examples (posting models).
Is noise in the background audio allowed (outdoors, street sounds)?
fluid topaz
<@823780286810161163>
its python 3.10+ issue, im using 3.9 to make it work
brittle wing
its python 3.10+ issue, im using 3.9 to make it work
so any fix for python 3.10+
proud elbow
so any fix for python 3.10+
there will be conflicting & broken packages
fluid topaz
yea i just downgraded to 3.9 rather than fixing those things
brittle wing
how downgrad
proud elbow
delete and recreate the venv w python 3.9, then install again
brittle wing
how do i do this help me please @tired kraken @acoustic scarab
acoustic scarab
how do i do this help me please <@678457649431314479> <@362876249032359939>
do what
exactly
brittle wing
like, can you guide me thru how to get the voice changer i seen on tiktok.
please...
acoustic scarab
like, can you guide me thru how to get the voice changer i seen on tiktok.
i've tagged you in #1159290666019471421 with several guides. Direct link to message - #1159290666019471421 message
tired kraken
Sorry I'm an AI image art helper not a Voice Changer helper.
acoustic scarab
(afaik all bots work only in #1159290666019471421 )
brittle wing
Sorry I'm an AI image art helper not a Voice Changer helper.
i didnt see the role, thats my fault
rare gobletBOT
Ayo? @brittle wing level 1 !!!
acoustic scarab
the voice changer itself called w-okada and has a dedicated help channel, https://discord.com/channels/1159260121998827560/1159290161683767298, feel free to ask anything about it there
violet heron
Raven
acoustic scarab
violet heron
like, can you guide me thru how to get the voice changer i seen on tiktok.
-realtime
azure marshBOT
-realtime
This interaction has expired, use the command
/guides realtimeif you wish to see it again.
acoustic scarab
And wrong channel <#1159290161683767298>
violet heron
im js downloding the rvc one
Good luck with that one
brittle wing
okay nvm
ts js crached
crashe
i got this 1
ts
this one, but its so shit
idk how to fix it
violet heron
this one, but its so shit
Change crepe to RVMPE
Change chunk to 192 iirc
Change audio to client
violet heron
oh cs idk the tiktok showed
This is why you don’t watch TikTok tutorials
violet heron
Set input to your mic
brittle wing
i want it to go thru my discord
violet heron
Set input to your mic
Set monitor to your headphones if you want to hear yourself
violet heron
i want it to go thru my discord
To use on discord you need a virtual audio cable
violet heron
the vb shit
Set output to it
brittle wing
outpit to vp input or vp out
violet heron
outpit to vp input or vp out
Output should be VB input iirc
violet heron
Yeah
brittle wing
then here should be?
violet heron
The cable output
jade hornet
hi, I select my mic in the app but i dont know why the voice isnt detected
cloud cradle
"the system cannot find the path specified" when i try to run gui_v1
lime sorrel
hi i need help
violet heron
"the system cannot find the path specified" when i try to run gui_v1
You run infer-web
lime sorrel
why does it throw this error, ive used RVC v2 "the disconnected colab"
brittle wing
why this dont work
@violet heron bro my gpu not even bad
why aint this running good./
:/
violet heron
why aint this running good./
Gain is too high
lime sorrel
am i doing something really dumb?
brittle wing
Gain is too high
okay here
rare gobletBOT
Ayo? @brittle wing level 2 !!!
cloud cradle
""PS E:\RVC1006Nvidia> & C:/Users/fazbe/AppData/Local/Microsoft/WindowsApps/python3.11.exe e:/RVC1006Nvidia/infer-web.py
Traceback (most recent call last):
File "e:\RVC1006Nvidia\infer-web.py", line 3, in <module>
from dotenv import load_dotenv
ModuleNotFoundError: No module named 'dotenv'
PS E:\RVC1006Nvidia> "" is a error i get?
tried pip installing it but no luck
rare gobletBOT
Ayo? @cloud cradle level 1 !!!
violet heron
how do i add like custom ones, like famous youtubers and shi
Click edit
Scroll down until you find a empty model slot
Upload the models .pth and .index file
You can find voices in #1175430844685484042 or the website weights.gg
lime sorrel
why does it throw this error despite me using the rvc v2 colab?
violet heron
why does it throw this error despite me using the rvc v2 colab?
Did bro try to select a training file for inference
lime sorrel
Did bro try to select a training file for inference
i thought the tutorial told me to add the .pth files into a zip
half cove
i thought the tutorial told me to add the .pth files into a zip
Use model name.pth
half cove
I'm gunna kill my autocorrect
lime sorrel
oh
half cove
D and g pth are for training
lime sorrel
D and g pth are for training
i dont see the model name.path in the rvcdisconnect
is it not inside of that?
half cove
Screenshot
lime sorrel
Bro got the wrong file
yes
lime sorrel
Screenshot
this is the whole entire folder
rare gobletBOT
Ayo? @lime sorrel level 4 !!!
lime sorrel
i hope im blind
violet heron
this is the whole entire folder
On the colab you used to train
Did you run export model weights cell
lime sorrel
smh im blind am i right
lime sorrel
Did you run export model weights cell
i think yeah
violet heron
On the colab you trained it on
half cove
violet heron
On the colab you trained it on
Go to your files
Find your way to the weights folder
That’s the file
lime sorrel
im so retarded
the weight ive exported, ive exported yesterday because i trained yesterday. now i cant find it because the website has restarted
idk if i lost the model ive exported but im pretty sure i did.
half cove
dawn vigil
I'm new to RVC, go-realtime-gui.bat and go-realtime-gui_new.bat seem to be in chinese? Is this right?
cloud cradle
I'm new to RVC, go-realtime-gui.bat and go-realtime-gui_new.bat seem to be in ch...
how did u get it to run?
dawn vigil
go-web.bat is in english, but I am just running the Mangio RVC Fork with the kalomaze fixes
rare gobletBOT
Ayo? @dawn vigil level 1 !!!
brittle wing
um
how do i add my
own
sound
@violet heron hey partner./
i may need a lil hand.
brittle wing
likeeeeeeeee
model
like ykk
how theres that anime, i wanan add a
youtuber orrr
someone
violet heron
brittle wing
dawn vigil
Do RVC models have a 'safetensor' equivalent. Or are there safety concerns to be aware of when downloading models? Is there a reliable site that scans models like civit does for SD?
spare shale
Day #100
ionic roost
Ok, I have a fairly unique problem that needs solving
I'm trying to make a decent model for a TTS
Getting clean audio isn't an issue, but the AI mimicks the TTS so well that it gets all the voice artifacts the TTS creates
So I would think that I somehow need to mix a human voice into the dataset that can hit the high and low notes
Has anyone else done something along these lines before, or have any ideas on how to help me?
I know of someone who did this before but he won't share his secret
forest sluice
@rigid light Can you run this command on mainline python path/to/mainline-rvc/infer/modules/train/extract/extract_f0_rmvpe.py 2 0 0 path/to/mainline-rvc/logs/test True
Shit's literally 1-1 port from mainline
Wait, I need to change True to False
forest sluice
@rigid light Replace FloatTensor with tensor.half (both lines) here https://github.com/IAHispano/Applio/blob/c36b57ce56c4108be6dd9e50ce5ad4258d08506d/lib/modules/infer/rmvpe.py#L112
cyan pilot
where and how can i make my own voice model yall
forest sluice
-guide
cyan pilot
ty 🫡
rare gobletBOT
Ayo? @cyan pilot level 1 !!!
cyan pilot
what do i need in my zip folder? just the ai voice right?
brittle wing
Does anyone know why the program does not recognize my gpu? what is the solution?
viscid moss
what do i need in my zip folder? just the ai voice right?
u need a .pth and ADDED___ .index file
brittle wing
i have a rx5600xt
rare gobletBOT
Ayo? @brittle wing level 1 !!!
brittle wing
spanish moddd te viii
viscid moss
i have a rx5600xt
#🔍│help-w-okada and send a screenshot
viscid moss
spanish moddd te viii
Holi
brittle wing
holaa
viscid moss
voice changer?
brittle wing
¿debo enviar screenshot?
brittle wing
Si, dejame ver una foto
ya lo subi al canal que me dijiste
viscid moss
https://docs.google.com/document/d/e/2PACX-1vQIwJ3MVidhgEaXwWFl0xpVonVOVfneaNVADd7-NMWFgPIsfWWhG8NNqzQMsXDIOGlBIfxscoIm2_6I/pub
Try searching in this doc ur issue
rare gobletBOT
Ayo? @brittle wing level 1 !!!
heavy shuttle
does anyone know why the google colab rvc v2 crepe + persistent storage / autosave hasnt been updated with the new pretrains?
im tried of uploading them myself and waiting for it to load.
viscid moss
does anyone know why the google colab rvc v2 crepe + persistent storage / autosa...
bro what
that colab it outdated
use the rvcDisconnected
heavy shuttle
that colab it outdated
well the disconnected one sucks training is way faster on the rvc crepe one
with the same dataset it takes 20 seconds per epoch
and on the disconnected it takes 40 - 50 seconds per epoch
heavy shuttle
<:idk2:632256425489465384>
but disconnected have the new pretrains
but the training takes LONGER than it needs to be 💀
alright give me a sec
heavy shuttle
<:idk2:632256425489465384>
but disconnected have the new pretrains
i dont know if you can access it
oh no it can be accessed lol
modern stag
can someone help with an error
"/content
FileNotFoundError Traceback (most recent call last)
<ipython-input-8-71023684ee54> in <cell line: 16>()
31 get_ipython().system('rm /content/srv.txt /content/srvr.txt')
32 elif tunnel == "gradio":
---> 33 for line in fileinput.FileInput(f'Easier{weeee}.py', inplace=True):
34 if line.strip() == 'app.queue(concurrency_count=511, max_size=1022).launch(quiet=True)':
35 # replace the line with the edited version
1 frames
/usr/lib/python3.10/fileinput.py in _readline(self)
355 pass
356 # The next few lines may raise OSError
--> 357 os.rename(self._filename, self._backupfilename)
358 self._file = open(self._backupfilename, self._mode,
359 encoding=encoding, errors=self._errors)
FileNotFoundError: [Errno 2] No such file or directory: 'EasierGUI.py' -> 'EasierGUI.py.bak'"
forest sluice
"/content
----------------------------------------------------------------------...
Colab link?
modern stag
nevermind I restarted the runtime thing and it fixed
heavy shuttle
<:idk2:632256425489465384>
but disconnected have the new pretrains
and then i cant even inference
to test my model
forest sluice
Look at the All-In-One Documentation
stray bloom
PYTHON:3.10.11 (tags/v3.10.11:7d4cc5a, Apr 5 2023, 00:38:17) [MSC v.1929 64 bit (AMD64)]
Activating the Voice Changer.
('Connection aborted.', ConnectionResetError(10054, 'An existing connection was forcibly closed by the remote host', None, 10054, None))
('Connection aborted.', ConnectionResetError(10054, 'An existing connection was forcibly closed by the remote host', None, 10054, None))
('Connection aborted.', ConnectionResetError(10054, 'An existing connection was forcibly closed by the remote host', None, 10054, None))
('Connection aborted.', ConnectionResetError(10054, 'An existing connection was forcibly closed by the remote host', None, 10054, None))
('Connection aborted.', ConnectionResetError(10054, 'An existing connection was forcibly closed by the remote host', None, 10054, None))
('Connection aborted.', ConnectionResetError(10054, 'An existing connection was forcibly closed by the remote host', None, 10054, None))
whhen i run start_http
im also running onnxgpu
because my pc has an nvidia gpu
anyone know how to fix?
azure marshBOT
RVC Guides (How to Make AI Cover)
scarlet wharf
its getting stuck here
dang i cant send link
its stuck at the step MMVC_SocketIOAPP initializing... done.
it doesnt open the little menu
potent edge
can someone help me?
i have this folder, but i dont know what to use it for
it has a file that ends with ".pth"
?
earnest jasper
modern oxide
wassup