theres a button at the bottom of training to export the lowest loss of the model
#✨│ai-help
serene solstice
sc? I dont see the option
serene solstice
this is mine rn
snow hazel
Fixed it!
serene solstice
maybe your save frequency is too high idk
serene solstice
bump
serene solstice
if rank == 0 and not hps.if_stop_on_fit:
if global_step % hps.train.log_interval == 0:
lr = optim_g.param_groups[0]["lr"]
logger.info(
"Train Epoch: {} [{:.0f}%]".format(
epoch, 100.0 * batch_idx / len(train_loader)
)
)
# Amor For Tensorboard display
if loss_mel > 75:
loss_mel = 75
if loss_kl > 9:
loss_kl = 9
logger.info([global_step, lr])
logger.info(
f"[loss_disc={loss_disc:.3f}] | [loss_gen={loss_gen:.3f}] | [loss_fm={loss_fm:.3f}] | [loss_mel={loss_mel:.3f}] | [loss_kl={loss_kl:.3f}]"
)
found the relevant code.
no clue what rank is though
serene solstice
rank is the gpu index it seems
so the true question on how often is prints is if global_step % hps.train.log_interval == 0: not rank
"log_interval" 💀 im smart
ok the log interval is set in the configs
every 200 steps
should be pretty simple to increase the frequency of
modern surge
Hello, who has that colab error?
It seems like I'm painted on the wall because no one responds to me or it seems like I'm the only one who has this happen
fallow glacier
It seems like I'm painted on the wall because no one responds to me or it seems ...
or that no one who knows how to fix it has seen it
viscid moss
https://tenor.com/view/anya-forger-anya-spy-x-family-anime-salute-salute-gif-okey-dokey-gif-25700022
viscid moss
It seems like I'm painted on the wall because no one responds to me or it seems ...
Bro, CUDA problems was fixed by me, Hina and Grd. Check the new notebooks
viscid moss
It seems like I'm painted on the wall because no one responds to me or it seems ...
-colab
mighty parcel
it's just noisy, also, looks like UVR vocals
fallow glacier
it's just noisy, also, looks like UVR vocals
That's because it is. There's no availabile stems
split shadow
Hey guys, I'm getting a weird error I don't understand when running the last cell in so-vits-svc 4.0. can anybody help me?
finite galleon
Hey guys, I'm getting a weird error I don't understand when running the last cel...
Why sovits?
split shadow
not sure, this is my first time trying to learn this. not sure if im doing it properly lol.
any guidance would be appreciated my friend!
split shadow
Why sovits?
im looking at the rvc links colelction channel, how do I know which one I should use?
rare gobletBOT
Ayo? @split shadow level 1 !!! 
azure marshBOT
RVC Guides (How to Make AI Cover)
glad zealot
Bro, CUDA problems was fixed by me, Hina and Grd. Check the new notebooks
They probably have an older copy of it.. that's why I'm planning on switching to codes hosted on pastebin so when i update stuff I just need to update the pastebin and all copy will be updated
serene solstice
is any uvr model particularly good at removing harmonies?
glad zealot
Karaoke2 or something
serene solstice
tried was meh. wish more songs had official acapellas 😭
glad zealot
You could try looking for a clip or video of them being interviewed or something
serene solstice
this is for inferencing
my model is getting a voice crack on the harmony and pitch shifting isnt really helping too much
proud elbow
is any uvr model particularly good at removing harmonies?
try UVR 5HP or 6HP, though can be hit or miss
sand iris
Which colab notebook are you using?
The error you mentioned doesn't really make any sense, since it seems to complain about a missing version that hasn't been used in a really long while
serene solstice
tried HP_2. these names are confusing lol. will give those a shot
god VR architecture is outputting some ear bleeding digital artifacting on some models
idk if it's cause linux or cause im using a short sample
finite galleon
my model is getting a voice crack on the harmony and pitch shifting isnt really ...
Karaoke 2 model should do the trick
serene solstice
karaoke 2 did not do the trick
finite galleon
karaoke 2 did not do the trick
Uvr BVE?
serene solstice
uvr idk what bve is
glad zealot
Most of the time karaoke 2 gives no change or just sounds muffled for me
serene solstice
god VR architecture is outputting some ear bleeding digital artifacting on some ...
gonna try on the full song instead of just a sample I guess
finite galleon
Most of the time karaoke 2 gives no change or just sounds muffled for me
Muffled when generated but after inferred it usually becomes unoticeable
glad zealot
Muffled when generated but after inferred it usually becomes unoticeable
No I use it for training so not usable
finite galleon
No I use it for training so not usable
Now thats a different story
glad zealot
Yeeee
finite galleon
I just remove every harmony i see
serene solstice
im just tryna get my model to not have a voice crack on that part of the song
brittle wing
Is the voice changer still banned for Google Colab?
glad zealot
Is the voice changer still banned for Google Colab?
Yesnt
brittle wing
Whats the diffrence?
serene solstice
the new one isnt banned
glad zealot
Whats the diffrence?
The new one works that's the difference lol
brittle wing
lol
proud elbow
No I use it for training so not usable
that's why I barely use soundtracks as part of training data, only for making covers
glad zealot
that's why I barely use soundtracks as part of training data, only for making co...
I use streams now.. well vods.. just chatting vods mostly so I don't have to do stuff
brittle wing
Will the vc work better with Colab Pro?
Or is it almost the same
I saw you get a better cpu with it
serene solstice
hi bot
brittle wing
Aight
sand iris
Squishy got it faster than I did lol
serene solstice
shot through the heart, niah's to blame
glad zealot
Wuut
proud elbow
I use streams now.. well vods.. just chatting vods mostly so I don't have to do ...
talking and karaoke streams, exactly (most have no or less problematic backing vocals)
glad zealot
talking and karaoke streams, exactly (most have no or less problematic backing v...
Ye since it's a stream I just remove background audio and use it
The manually check if there's some left to remove manually
crisp goblet
Quickest way to make a voice model?
rare gobletBOT
Ayo? @crisp goblet level 1 !!!
crisp goblet
i mean dataset
proud elbow
Quickest way to make a voice model?
quality matters much more, go check and learn some #1159513888199540817
crisp goblet
I meant dataset
whats a way I can make a high quality model on my device
I have the hardware
if someone can send me their confiuration or a link to an automatic vocal splitter that would rlly help me
finite galleon
I do uh MDX23C > VitLarge > VocFT > FoxJoy Dereverb > DeEcho Normal > Denoise which all can be found on mvsep.com
Or UVR if you want to do it locally
crisp goblet
Does that cut out silence and make seperate audio files?
I'm trying to voice clone my homie
thin stump
What's vitlarge? I'm assuming that's only for making models?
crisp goblet
I'm just gonna try what I have
silver anvil
sand iris
compromised acc spamming links
harsh scarabBOT
You're placed in Cover Generator Queue. Current position: 1
twilit ferry
hey my applio keeps saying i dont got a module names 'tensorboard'
rare gobletBOT
Ayo? @twilit ferry level 1 !!!
azure marshBOT
RVC Guides (How to Make AI Cover)
viscid moss
They probably have an older copy of it.. that's why I'm planning on switching to...
ya ik that's why I overwrote the changes on the previous notebook.
glad zealot
ya ik that's why I overwrote the changes on the previous notebook.
use pastebin method :D
viscid moss
use pastebin method :D
Like Rentry? 
distant wagon
Uh, how to do this?
rare gobletBOT
Ayo? @distant wagon level 1 !!!
distant wagon
glad zealot
on the part that says "url" you put the huggingface link of the model you want
old oracle
-colab
rare gobletBOT
Ayo? @old oracle level 2 !!!
distant wagon
on the part that says "url" you put the huggingface link of the model you want
I am very confused
glad zealot
just put the huggingface link of the model
distant wagon
qwq??
glad zealot
yes
serene solstice
Are there any good methods at manipulating portions of an acapella to make a model handle inferencing specific high notes better? I've tried adjusting pitch of problematic segments and that sometimes works? But it can also make it sound wonky. I mainly use Audacity but I can look into whatever else might work better for any given method.
glad zealot
you can manually pitch down the parts where its too high but idk what program that is
proper shale
melodyne?
glad zealot
no idea
proper shale
melodyne
serene solstice
And that would differ from adjusting the pitch in audacity?
glad zealot
you can do specific parts.. dont know if thats possible with audacity
proper shale
And that would differ from adjusting the pitch in audacity?
yeah
its kinda like autotune
serene solstice
You can select segments in Audacity and apply a gradual pitch change but like I said it kinda turns out meh. Might give melodyne a shot then.
glass oar
How can we use this AI for a group of people? For example, if I want to change the voice in a song that is sung by 4 people(A, B, C, D) into 4 ai voices(E, F, G, H). Should I separate the parts and apply ai one by one or is there a easier way? And, if I have to separate the parts, how can I do it?
proud elbow
How can we use this AI for a group of people? For example, if I want to change t...
one model for one person, also should avoid duets/group vocal songs
glass oar
Then is there a easy way to separate it? Like an AI or a program?
tame mica
manual
tidal shadow
anyone have 2ne1 Models?
upbeat pine
Can anybody help me??? Every time I use Google colab, when doing the last function it appears: File "/content/AICoverGen/src/main.py", line 344, in <module>
raise Exception(f'The folder {os.path.join(rvc_models_dir, rvc_dirname)} does not exist.')
Exception: The folder /content/AICoverGen/rvc_models/Sakura does not exist.
1
distant wagon
Enter it here? @glad zealot
brittle wing
if i have a dataset thats 27 min how many epochs should it be for training?
lament vessel
Question would like to train a voice model. Online and that has a ui. (im terribly bad with typing and redirecting stuff T-T and didnt even check the links properly cause lazyyy💀)
twilit ferry
my applio keeps saying its missing the tensorboard and i cant do anything, can anyone help?
golden anvil
i started the client it runs fine but when i hit the start button it says:
[Voice Changer] VC PROCESSING EXCEPTION!!! CUDA error: no kernel image is available for execution on the device
CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
For debugging consider passing CUDA_LAUNCH_BLOCKING=1.
Compile with TORCH_USE_CUDA_DSA to enable device-side assertions.
Traceback (most recent call last):
File "voice_changer\VoiceChangerV2.py", line 229, in on_request
File "voice_changer\RVC\RVCr2.py", line 197, in inference
File "voice_changer\RVC\pipeline\Pipeline.py", line 161, in exec
sand iris
i started the client it runs fine but when i hit the start button it says:
[Voic...
What's your GPU
golden anvil
nvidia
rare gobletBOT
Ayo? @golden anvil level 1 !!!
golden anvil
GTX710
sand iris
Yup, too old and underpowered to work unfortunately
golden anvil
alrighty
sand iris
you can check out the pins in #🔍│help-w-okada , there's a way to run this online on google servers
It's pretty much the exact same thing with extra latency from the network
golden anvil
ok tysm
indigo topaz
is covergen working?
spiral tapir
Hi there, it keeps giving me this error when I try to process, what does it mean?
runtime\python.exe trainset_preprocess_pipeline_print.py C:\Users\Pc\Music\Nyarls\Dataset\Nyarls.wav 40000 8 D:\Mangio RVC\Mangio RVC 2\Mangio-RVC-v23.7.0/logs/Nyarlathotep False
Traceback (most recent call last):
File "D:\Mangio RVC\Mangio RVC 2\Mangio-RVC-v23.7.0\trainset_preprocess_pipeline_print.py", line 25, in <module>
f = open("%s/preprocess.log" % exp_dir, "a+")
FileNotFoundError: [Errno 2] No such file or directory: 'D:\Mangio/preprocess.log'
rare gobletBOT
Ayo? @spiral tapir level 1 !!!
proud elbow
Hi there, it keeps giving me this error when I try to process, what does it mean...
don't use spaces in the directory names (MangioRVC instead of Mangio RVC)
tacit jetty
don't use spaces in the directory names (`MangioRVC` instead of `Mangio RVC`)
How long should i let the model train
reef meadow
Hey all, I’m new to discord and fairly new to rvc. Previously I’ve been replay to make covers but I want to start start training my own models. I’ve downloaded rvc_infer on my Mac and have an audio file ready but am clueless on how to proceed. Any help is appreciated
old oracle
-colab
spiral tapir
don't use spaces in the directory names (`MangioRVC` instead of `Mangio RVC`)
Ah that fixed it thank you, but now I'm stuck on another error sorry to bother you again
start preprocess
['trainset_preprocess_pipeline_print.py', 'C:\Users\Pc\Music\Nyarls\Dataset\Nyarls.wav', '40000', '8', 'D:\MangioRVC\MangioRVC\Mangio-RVC-v23.7.0/logs/Nyarls', 'False']
Fail. Traceback (most recent call last):
File "D:\MangioRVC\MangioRVC\Mangio-RVC-v23.7.0\trainset_preprocess_pipeline_print.py", line 118, in pipeline_mp_inp_dir
for idx, name in enumerate(sorted(list(os.listdir(inp_root))))
NotADirectoryError: [WinError 267] The directory name is invalid: 'C:\Users\Pc\Music\Nyarls\Dataset\Nyarls.wav'
end preprocess
why is it invalid I don't get it
cold surge
hello, could i get a guide on how to install / use rvc model please?
proud elbow
Ah that fixed it thank you, but now I'm stuck on another error sorry to bother y...
double check if the path name actually exists
sand iris
Oh no you're supposed to put a folder
Not a path to a file. Yeah it's annoying, but the fix is just to put that single .wav file in a folder
spiral tapir
a folder?
proud elbow
yeah, a folder path that contains at least a .wav file (i.e. probably C:\Users\Pc\Music\Nyarls\Dataset)
rare gobletBOT
Ayo? @proud elbow level 12 !!!
sand iris
yeah, a folder path that contains at least a .wav file (i.e. probably `C:\Users\...
^ Assuming there's only that .wav file in there
spiral tapir
^ Assuming there's only that .wav file in there
Ah Thank you that fixed it
peak imp
im curious im inputting a google link to download the huggingface model thing
does it bhave to be a huggingface link
some reason my model wont show now
nvm
reef meadow
Whenever I try to train the model I get this message start preprocess
['trainset_preprocess_pipeline_print.py', '/Users/jayturner/Desktop/Mangio-RVC-Fork-23.7.0/datasets/jay', '40000', '6', '/Users/jayturner/Desktop/Mangio-RVC-Fork-23.7.0/logs/mi-test', 'False']
end preprocess
rare gobletBOT
Ayo? @reef meadow level 1 !!!
proper shale
Whenever I try to train the model I get this message start preprocess
['trainset...
It's fine
proper shale
you're welcome!
reef meadow
Also, does this mean that it’s training the model?
proper shale
Also, does this mean that it’s training the model?
this only appears at the end of training
reef meadow
this only appears at the end of training
But it appeared immediately after clicking train model. I’ve never done this before but I was under the impression it takes some time lol. If it is done, the completed model should be in the weights folder?
proper shale
But it appeared immediately after clicking train model. I’ve never done this bef...
Well I think something went wrong
Did you feature extract?
reef meadow
Yes, I got this message ['extract_fO_print.py,
/Users/jayturner/Desktop/Ma
ngio-RVC-Fork-
23.7.0/logs/mi-test, '6,
'rmvpe', '64]
fofail-O-
tacit jetty
Is macos training supported <@876833920035999755> ?
<:nails:1159569314848972891>
what the frick does this mean
old oracle
-colab
proper shale
<@622856015444049937>
good question, im unsure
tacit jetty
How long does Export Model from Notebook to Drive
take
azure marshBOT
RVC Guides (How to Make AI Cover)
azure marshBOT
azure obsidian
https://huggingface.co/ROSES/adam/resolve/main/AdamElevenLabs.zip yall this my model and when i upload it to ilaria RVC it will fail
strange crane
E:\RVC\Mangio-RVC-v23.7.0\runtime\lib\site-packages\torch\autograd_init_.py:200: 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:337.)
Variable._execution_engine.run_backward( # Calls into the C++ engine to run the backward pass
Training time per epoch has increased by ~1 min (1m 22s -> 2m 30s). I'm guessing this is the cause as I don't remember seeing this message before. Anyone know how to fix this?
keen pivot
How i make AI Songs
loud kelp
anyone has a link where i can make ai songs?
loud kelp
does applio work?
thin stump
Yes, people still use it. I couldn't find the exact guide on it
loud kelp
where can i find a link for applio?
thin stump
where can i find a link for applio?
https://discord.com/channels/1159260121998827560/1159618293594861579 My bad, I didn't add #1159513888199540817 to my channel list
polar axle
How can I make my own voice into my own voice model for Replay?
rare gobletBOT
Ayo? @polar axle level 2 !!!
steady ruin
Yes, I got this message ['extract_fO_print.py,
/Users/jayturner/Desktop/Ma
ngio-...
If you get a message in the console that says something like "CUDA out of memory" then it definitely didn't train (I learned this firsthand). Your GPU might be too weak to train, in which case you should train with a Colab or on HF, or something else might've gone wrong, which I unfortunately don't have a solution to.
mental spruce
what version on cuda do i need ?
so i dont need cuda ?
?
ok i have
pip and pythorch
and python
got the voice changer set
wdym ?
meager terrace
Hey, anyway to find earliest pth if I resume training on RVCDisconnected?
young halo
How can I get the link to the model?
steady ruin
-colab
steady ruin
-overtrain
azure marshBOT
All-In-One Guide on how to make a good model
This guide explains how the D and G files works and much more: https://rentry.org/RVC_making-models
Credits: LUSBERT 
Automated Overtraining Detection (AOD)
Will be available soon in #1159513888199540817
Credits: grvyscale
young halo
Thank you!
lavish meteor
Guys, i keep getting this error, could you help me out?
\Mangio-RVC-v23.7.0/logs/nefi/trained_IVF1085_Flat_nprobe_1_nefi_v1.index for writing: No such file or directory
lavish meteor
Guys, i keep getting this error, could you help me out?
\Mangio-RVC-v23.7.0/log...
i am getting this on train future index part.
reef meadow
If you get a message in the console that says something like "CUDA out of memory...
Ty, I’ll check it out. Thought my m3 Mac would be strong enough to train. Also I thought google collab was shut down, and what is hf?
meager terrace
okay, got it, very clear, thank you ❤️
lavish meteor
Could you guys help me too?
steady ruin
Ty, I’ll check it out. Thought my m3 Mac would be strong enough to train. Also I...
oh there are some GUI-less colabs that are good to use
because they don't use an external WebUI
lavish meteor
\Mangio-RVC-v23.7.0/logs/nefi/trained_IVF1085_Flat_nprobe_1_nefi_v1.index for writing: No such file or directory
rich yoke
i am beginning to tweak out trying to get this to work.
the install is literally
IV. Install Prerequisites: Run the "install-requirements.bat" script to install the necessary dependencies.
what is the install requirements.bat
where do i get that
neither of my launch tensorboards work
if mangio so smart then what doesn't it just start when you start training or just have a 'launch tensorboard' in the interface itself. why are we hiding it
brittle wing
is this where we ask questions about the voice models?
azure marshBOT
RVC Guides (How to Make AI Cover)
thin stump
Is there a way to remove some of the vocal bleeds on the instrumentals? I used MDX23C-InstVoc and then denoise aggressive on instrumentals. I still hear some of it
proper shale
is this where we ask questions about the voice models?
yeah, about rvc in general
proper shale
Is there a way to remove some of the vocal bleeds on the instrumentals? I used M...
Maybe try the Inst_HQ MDX-Net stuff?
proper shale
Maybe try the Inst_HQ MDX-Net stuff?
nope they're just Inst models
not Inst_HQ
Anyways try them out
thin stump
I forgot I can combine models, i'll be really thorough with it. The vocals extracted were perfect already
modern surge
Colab's bug has already been fixed? I've been asking for help or a response for about 4-5 days
😭
modern surge
Which colab notebook are you using?
The error you mentioned doesn't really make ...
no ui colab
-colab
soft tiger
whats a software where i can make vocal mp3 files into the ai models voice
azure marshBOT
soft tiger
Would Kim vocal 2 be the same as 1?
ik you aint a mod but do you know?
proper shale
whats a software where i can make vocal mp3 files into the ai models voice
...rvc?
proper shale
u can use the colabs or the local stuff if u have a good pc
rare gobletBOT
Ayo? @soft tiger level 2 !!!
soft tiger
u can use the colabs or the local stuff if u have a good pc
uhhh
which ever is the easiest
proper shale
I'd recommend reading the guides n stuff
proper shale
which ever is the easiest
Well let me be clear with you, this stuff ain't simple
But the guides will help
soft tiger
Well let me be clear with you, this stuff ain't simple
i had this one software on my pc that was like frag and drop lol
proper shale
If you just read carefully n shit it'll work
soft tiger
but that was months ago lol
proper shale
-colab
distant yarrow
would yall say RVC is better with the realtime AI voices than Okada?
proper shale
RVC Disconnected is a good one
proper shale
would yall say RVC is better with the realtime AI voices than Okada?
...okada uses RVC?
distant yarrow
...okada uses RVC?
...ah
soft tiger
uh uh
proper shale
i mean there is a realtime voice changer built in to rvc but it's a bit clunky to use
distant yarrow
I see, so it must just have been a crappy voice model I was using then
soft tiger
is there any links for the local softwarte you was saying
proper shale
is there any links for the local softwarte you was saying
-local
azure marshBOT
Suggestions for @soft tiger
rich yoke
I'm 208 epochs in and tensor is still showing improvement. Thanks for your help again!
soft tiger
Suggestions for <@407013052001681411>
how do i run original rvc
idk what half of this is
rich yoke
so i should keep training a little bit after g/total stops going down then look and see if g/mel goes down even more near it and test/compare them?
im unsure of what would be considered a 'good' dataset, mines probably mediocre
yes
i noise gated and did my best ensemble uvr5 pass then like 4 other passes of noise, verb, then noise verb noise verb
snow hazel
Ok so on audacity it says that my song is 44100 Hz but I get an error on mangio.
rich yoke
i just realized that i left this giant patch of nothing in my dataset when everything past like 15mins was supposed to be cropped entirely
do you have any guides on euthanasia
snow hazel
snow hazel
I need help
rich yoke
this is a bit over 10 seconds zoomed in
what defines quality, how can you tell by looks?
i see
so you want many defined features
with negative space like that
i imagine it's because my original source is a singer who uses a lot of verb and echo
is there an izotope rx10 and audacity guide
for using their features to clean
snow hazel
I need help
trim patrol
snow hazel
dang
floral tusk
How do I use models?
marble shuttle
I'm going to start asking questions here. I'm looking to find out where I can get the WordBand Bot.
rare gobletBOT
Ayo? @rich yoke level 4 !!!
desert bison
Hi, im really new to all this, can someone help?
i downloaded a Morty Smith rvc but it does not work properly
weak briar
what are the optimal settings to remove artifacts in medium tempo songs?
molten pecan
I'm going to start asking questions here. I'm looking to find out where I can ge...
for your server?
molten pecan
i downloaded a Morty Smith rvc but it does not work properly
RVC is the name of the software for making the AI covers. Morty's voice is what the <voice model> is about
so, you want to do inference with that model but it's not working properly. what's the issue exactly?
molten pecan
what are the optimal settings to remove artifacts in medium tempo songs?
well the artifacting issue could be due to a few issues
-Audio is overall low-quality
-Model was overall poorly trained
-There are overlapping voices
-There is reverb or other sound effects
-Or there is noise
azure marshBOT
-Audio is overall low-quality
-Model was overall poorly trained
-There are overl...
- Perfecting Audio Isolation on Low-End Rigs: A Practical Guide, by Litsa The Dancer and Faze Masta Google Docs
- Gathering and Isolating Audio, by SCRFilms Google Docs
- How to make a good model All-In-One guide, by LUSBERT Rentry
- Vocal Mixing Tutorial, by Roomie YouTube
molten pecan
but try reducing the search feature ratio
desert bison
so, you want to do inference with that model but it's not working properly. what...
it seems like it just only muffles my sound when i listen back to it
rare gobletBOT
Ayo? @desert bison level 1 !!!
molten pecan
it seems like it just only muffles my sound when i listen back to it
muffles?
wdym exactly?
sharp veldt
Im trying to start up rvc gui but it gives me a virus warning is it safe orrrr???
any other programs better??
weak briar
-Audio is overall low-quality
-Model was overall poorly trained
-There are overl...
hm then it's probably exterraneous sound effects,
molten pecan
Im trying to start up rvc gui but it gives me a virus warning is it safe orrrr??...
yeah it usually does that. it's virus free, don't worry
weak briar
but try reducing the search feature ratio
but thanks!
I've never touched search feature
what is it, exactly?
molten pecan
but thanks!
yw! make sure your audio is in WAV or FLAC format too, not a lossy format like MP3
sharp veldt
And i have to do the full 5gb of file extraction right??
desert bison
i thought i can send here the audio file, is there any way to send here?
molten pecan
what is it, exactly?
This affects how much the of the .index file will be applied to the audio after inferencing. The higher the value, the more accent and speech manner of the voice model will be applied to the output audio.
For example, if the model's dataset has background noise, reducing this value will lessen that noise in the output.
Lowering this value can fix artifacting issues on some models.
Lowering it to 0 won't make the output not sound like the model, you can lower it as much as you want.
Here's more info on the settings https://rentry.org/ilarvc_inf_guide/#settings-explanation
my own copypasta :b
weak briar
This affects how much the of the .index file will be applied to the audio after ...
oh! that's awesome <3
molten pecan
And i have to do the full 5gb of file extraction right??
you mean unzipping it? yeah you should
sharp veldt
okay thanks
molten pecan
oh! that's awesome <3
yw !!!
molten pecan
i thought i can send here the audio file, is there any way to send here?
you can send it on my dms
weak briar
is there a place to post finalized songs (just curious, I'm actually in the wrong channel now)
molten pecan
i'll take a look at it
desert bison
alright, one sec ^^
molten pecan
is there a place to post finalized songs (just curious, I'm actually in the wron...
#1159290752195633273 youtube links only!
weak briar
<#1159290752195633273> youtube links only!
understood
rare gobletBOT
Ayo? @weak briar level 1 !!!
molten pecan
understood
have fun ^^
desert bison
i want to animate a different outcome or much like how Morty could react to Rick Primes death, it was already brutal but i muffled the actual audio on that clip and it's terrifying
desert bison
it's fine ^^
molten pecan
yeah so, the output sounds much more lower compared to his voice
did you try turning the pitch up?
in RVC
timid vessel
hey is orc here?
desert bison
yes, but it doesn't seem like it's doing anything
timid vessel
i was researching on reddit and saw someone named orc posting about okuda vc 😄
molten pecan
yes, but it doesn't seem like it's doing anything
what version of RVC are you using?
Applio?
desert bison
ooof, id on't know, where can i check that? 
molten pecan
ooof, id on't know, where can i check that? <:nails:1159569314848972891>
where did you download RVC from?
github? Hugging Face?
when I say version I mean fork. Do you know what that is?
desert bison
nope
rare gobletBOT
Ayo? @desert bison level 2 !!!
molten pecan
ohhh so you are using the voice changer
timid vessel
julia i was wondering if you could please help me with w okada and increasing the power in passing?
molten pecan
is that correct?
desert bison
ohhh so you are using the voice changer
yes yes
timid vessel
ive tried tweaking everything and watching youtube guides but my friends say it's still glitchy and has artifacts
molten pecan
julia i was wondering if you could please help me with w okada and increasing th...
I don't know what you mean by that. I also don't use w-okada so I can't help you. If you need help ask on #🔍│help-w-okada
timid vessel
ok ty
molten pecan
yes yes
ask on #🔍│help-w-okada I don't know much about the voice changer
desert bison
so there is two different apps?
molten pecan
so there is two different apps?
let's explain. RVC is the name of the AI software itself. out of that software you can make two tools. one is for converting the voice using audio files, and the other for changing your voice in real-time through the mic as you speak
rancid pine
Is there a tutorial on how to operate Tensorboard?
desert bison
well, im gonna ask then on the help-w-okada, thanks for your help! ^-^
rancid pine
Is there a tutorial on how to operate Tensorboard?
As in, I'm trying to launch it
molten pecan
Is there a tutorial on how to operate Tensorboard?
desert bison
there's plenty of info there
but first i will take a look at this x3
molten pecan
but first i will take a look at this x3
alright. have fun!
desert bison
thank yoouu :3
molten pecan
yw!
rancid pine
where? in local RVC?
Local RVC
rare gobletBOT
Ayo? @rancid pine level 1 !!!
molten pecan
Local RVC
I think this guide is outdated https://docs.aihub.wtf/guide-to-create-a-model/tensorboard-rvc
shit we need to make a new guide on launching the tensorboard 😭
@half cove wanna help me out on dat?
rancid pine
Ah damn
molten pecan
is there an up-to-date tensorboard launch guide?
rancid pine
I tried to follow a YouTube video, but I kept running into problems with Python using VSCode
sharp veldt
how do i get it to not like tweak out so much??
rare gobletBOT
Ayo? @sharp veldt level 1 !!!
azure marshBOT
RVC Guides (How to Make AI Cover)
steel cypress
hello! Guys, I need your help.
I currently created the voice using diff/rvc.
However, a problem arises when using this to extract songs.
AttributeError: 'NoneType' object has no attribute 'dtype'
Does anyone know why this happens? I tried almost every solution on github, but it doesn't work.
And how can I extract this into one file?
proper shale
hello! Guys, I need your help.
I currently created the voice using diff/rvc.
How...
diff/rvc?
like I'm confused here, what's that
steel cypress
OH sorry that is
RVC-Project / "Retrieval-based-Voice-Conversion-WebUI"
https://github.com/RVC-Project/Retrieval-based-Voice-Conversion-WebUI
rare gobletBOT
Ayo? @steel cypress level 1 !!!
trim cipher
Hello i need help on w-okada
w-okada voice changer is super lagging on my laptop and I am using
This is the specs of my laptop
4 Terabytes Ssd
32GB ram
Nvidia GeForce RTX 3080 16GB
Help I need help on why I am still getting a lag on the voice changer
azure marshBOT
RVC Guides (How to Make AI Cover)
lament vessel
Question would like to train a voice model. Online and that has a ui. (im terribly bad with typing and redirecting stuff T-T and didnt even check the links properly cause lazyyy💀)
spiral tapir
how long does it usually take to process the data? mine has been running for over 3500 seconds now
they say it usually takes less than 10 minutes but it's been almost an hour now, I'm wondering if there's an error or I did something wrong
brittle wing
yeah, about rvc in general
ive been making some ai covers for the fun of it and realised that they can't hold notes that well, is that a known problem or am i doing something wrong?
proven acorn
Is SVC no longer usable?
proven acorn
Are RVC1 models compatible with RVC2?
nvm
what does
Select Speaker/Singer ID:
mean
distant wagon
help
rare gobletBOT
Ayo? @distant wagon level 2 !!!
molten pecan
what does
Select Speaker/Singer ID:
mean
Is an unfinished feature in RVC that doesn't work.
The idea was to be able to store multiple voices within a single [.pth] file and access them by their assigned Speaker ID (SID). It most likely will remain unusable until much later on in the future
proven acorn
Really?
Well I wanted to use this
https://www.weights.gg/models/clm738u091vhhcctcgu4lcejj
molten pecan
what next<:boohooh:1176674698629750975>
I think you forgot to delete the text in path file
proven acorn
I assumed I needed that feature
distant wagon
I think you forgot to delete the text in path file
uh can you tell me where it is
molten pecan
uh can you tell me where it is
Where it says "enter the path of the audio folder"
molten pecan
Well I wanted to use this
https://www.weights.gg/models/clm738u091vhhcctcgu4lcej...
Maybe the voices are divided by their each .pth file. Are they?
molten pecan
Yes
proven acorn
Maybe the voices are divided by their each .pth file. Are they?
No
distant wagon
delete all?
proven acorn
It's just model.pth, model.index and some json file
It says that it includes all the 6 characters
but I have no idea how to access the other 5
molten pecan
delete all?
Yes. I think there are two of them in the page. Delete both text, unless you have already pasted a path file
distant wagon
ok
molten pecan
but I have no idea how to access the other 5
Me neither. I thought training multiple voices at once will make a bad sounding model
molten pecan
It just sounds like Twilight Sparkle
Maybe try using the speaker id feature
proven acorn
ok
molten pecan
Maybe now it works and I wasn't aware of it 😭
molten pecan
You are not supposed to paste a link to a website there
distant wagon
ohh
proven acorn
and i just looked at the json file and it's using a file from another model
i think
distant wagon
molten pecan
ohh
There you are supposed to paste the <path> to a <folder/file> from your computer
Is there a tab to download voice models?
Inside the app
lament vigil
HEROー🦸
molten pecan
uhhhhhhhhhhh
Is there a tab called Resources?
Are there any other tabs?
Can you scroll up and send a screenshot?
distant wagon
molten pecan
Congrats!
molten pecan
Now go back to the main tab, click Refresh, select your model, and you are good to go
distant wagon
distant wagon
I pressed it once before I used it
molten pecan
I pressed it once before I used it
Do this
molten pecan
Now go back to the main tab, click Refresh, select your model, and you are good ...
.
rare gobletBOT
Ayo? @distant wagon level 3 !!!
molten pecan
When I said main tab I meant the Inference tab
molten pecan
Here, upper right corner
distant wagon
molten pecan
Left corner I meant. Sorry
molten pecan
The click the blue Refresh button
And to the left of it, select your model
distant wagon
https://github.com/RVC-Project/Retrieval-based-Voice-Conversion-WebUI
in here?
molten pecan
In here
distant wagon
oh
molten pecan
Also if you want me to see your text, ping me next time
It's not like I stay in this channel 24/7
distant wagon
It's not like I stay in this channel 24/7
then can I dm you ?
distant wagon
It's okay if you ping me
I just want to say that it might be more convenient if I dm
molten pecan
I just want to say that it might be more convenient if I dm
I would rather stay here
Feel free to ping me
distant wagon
Feel free to ping me
and...i cant find it
distant wagon
molten pecan
Congrats, you found it
distant wagon
molten pecan
You are in the Inference tab
Hit Refresh and then click "Choose your model" on the left
Do you see the name of your model?
distant wagon
yes
molten pecan
Good
Select it
And then below it, select your audio file
Change the settings if you want, and then hit Convert
And wait for it to load
molten pecan
Will it give me options?
Options to what?
distant wagon
Options to what?
it doesn't give me the option
molten pecan
it doesn't give me the option
The option to what?
distant wagon
The option to what?
uhh do i need to enter anything in there?
molten pecan
uhh do i need to enter anything in there?
There where?
distant wagon
There where?
"Choose your model"
molten pecan
"Choose your model"
You need to click the rectangle and it will unfold the list of downloaded models
Click it and select your model
rancid pine
I think this guide is outdated <https://docs.aihub.wtf/guide-to-create-a-model/t...
I can't do shit with it
Rip
molten pecan
I can't do shit with it
@proper shale can you help him how to download tensorboard locally?
distant wagon
Did you click the arrow?
yes
molten pecan
yes
Then I think your model is not download
Go to the Download Model tab and paste the link and click Download
distant wagon
ok
molten pecan
Then go back to Inference, click Refresh, and see if it's there
distant wagon
Then go back to Inference, click Refresh, and see if it's there
is work
rare gobletBOT
Ayo? @distant wagon level 4 !!!
distant wagon
molten pecan
do i need to delete it
I don't think so this time
brittle wing
how do we open rvc? someone told me it was the ''go-web'' file, is it true? if its true, what is supposed to happen after i click it?
distant wagon
I don't think so this time
oh ok
brittle wing
because it opens up a black screen with white text
rancid pine
how do we open rvc? someone told me it was the ''go-web'' file, is it true? if i...
It automatically opens up the local hosting link for me
distant wagon
I don't think so this time
brittle wing
should i delete mine and download whatever you downloaded
molten pecan
because it opens up a black screen with white text
Yeah it opens a terminal. After a minute it will open the app itself
brittle wing
Yeah it opens a terminal. After a minute it will open the app itself
so i gotta wait?
molten pecan
Yeah delete the text then
brittle wing
it doesnt look like anything will happen
distant wagon
Yeah delete the text then
this?
distant wagon
ok
brittle wing
also windows sent a warning so idk what to do
should i allow the thing or not
btw it sent me to a site so i guess it worked
but what about windows warning thiny
thingy
tame mica
now paste the path to your audio and the audio file (e.g c:/name/folder/folder/audioname.wav)
replace the audioname with the name of your audio and c:/name/folder/folder with your path
distant wagon
now paste the path to your audio and the audio file (e.g c:/name/folder/folder/a...
Is it from here?
tame mica
oh you are using huggingface?
molten pecan
now paste the path to your audio and the audio file (e.g c:/name/folder/folder/a...
They are using hugging face
distant wagon
oh you are using huggingface?
yes
molten pecan
yes
Use my guide https://rentry.org/ilarvc_inf_guide
distant wagon
Use my guide https://rentry.org/ilarvc_inf_guide
?????
distant wagon
Use my guide https://rentry.org/ilarvc_inf_guide
So the huggingface side is complete?
distant wagon
molten pecan
?????
The hugging face you are using is a little hard for you. Use Ilaria RVC, it's simpler. I sent you a guide for it too
brittle wing
hey, what happens if i close the black screen? will i still be able to use the site?
idk if its called a site tho since you all call it host
but you know what i mean
molten pecan
Julia i recommend using screenshots btw
Too late I thought he was using applio locally 😭
tame mica
<:boohooh:1176674698629750975>
what huggingface space are you using
molten pecan
what huggingface space are you using
RVC V2
tame mica
Too late I thought he was using applio locally 😭
on the guide
molten pecan
i've never heard of it and seems like you can't(?) fully use it so...
Seems to just be mainline RVC on HF
distant wagon
👀
tame mica
Seems to just be mainline RVC on HF
yea and u can't use inference either
so just use ilaria rvc imo
distant wagon
If hugging face is used, are there many steps later?
tame mica
no not really
just the usual download models > inference
with an actual audio dropdown
molten pecan
yea and u can't use inference either
What 😭😭
That's so weird
Why would they put it then?
If it's there then it's for a reason
distant wagon
with an actual audio dropdown
Can I continue to use it?
tame mica
Can I continue to use it?
well yea but i don't know if you can inference
share the space link
distant wagon
tame mica
yea
rancid pine
hey, what happens if i close the black screen? will i still be able to use the s...
No
The process will halt
twilit ferry
is there a way i can send photos to help see my issue
rancid pine
is there a way i can send photos to help see my issue
Click the + button
brittle wing
guys, i get this error when i want to train a model
https://prnt.sc/BzAX7nddCg8t
ideas?
rancid pine
guys, i get this error when i want to train a model
https://prnt.sc/BzAX7nddCg8t
Don't you have to wait until 'end preprocess'?
brittle wing
it never gets there
rare gobletBOT
Ayo? @brittle wing level 1 !!!
rancid pine
Hm
boreal snow
how do i train a model? a am really confused about all the different versions of RVC, i think i have voice changer client demo... I don't know it that matters or not though.
brittle wing
can someone join vc for faster fixing?
twilit ferry
is it supposed to take this long😭
rancid pine
is it supposed to take this long😭
In my personal experience, it depends on the site you use
rare gobletBOT
Ayo? @rancid pine level 2 !!!
twilit ferry
its the one applio opened
proud elbow
is it supposed to take this long😭
check the console window for the epoch progress, also you can run tensorboard to monitor the progress (the graph)
south carbon
Need help on what could be causing lag?
I configured the voice (And converted to onnx), and set the delay to 128, everything worked perfectly without lags or problems. The next day I started it and didn’t change anything, and now the voice began to lag.
proud elbow
try start the runtime option
proud elbow
runtime, not conda
south carbon
Need help on what could be causing lag?
I configured the voice (And converted to...
I restarted the PC, etc., all the voices work on these settings, but the one I need stopped, it lags very much
molten pecan
how do i train a model? a am really confused about all the different versions of...
I'll help you with that
RVC has different forks. A fork is a copy of a main GitHub project. It aims to make a different version of the project with improvements, changes and new features
glad zealot
RVC has different forks. A fork is a copy of a main GitHub project. It aims to m...
i think they meant the cuda stuff
molten pecan
The most popular forks are Mangio and Applio. And then there's what we call "Mainline RVC", which is the base RVC
molten pecan
i think they meant the cuda stuff
They downloaded w-okada 😭
You can't train on that, right?
glad zealot
nope but they have it..
molten pecan
nope but they have it..
Yeah but they want to learn how to train a model
glad zealot
the easiest to use (from what i can see) it mainline for training
since you just get it, unzip and run
molten pecan
since you just get it, unzip and run
Random question but can you change the hop length of crepe for training in mainline?
glad zealot
Random question but can you change the hop length of crepe for training in mainl...
as far as i know no
unless you edit the configs probably
molten pecan
Dang
boreal snow
@glad zealot where do i find the mainline version? i dont see it in the list of versions.
i think i meant to reply
oops
peak imp
In my personal experience, it depends on the site you use
can you send me another site
this pink one im using is kinda weird
it keeps giving errors
obtuse dew
why can't I send pictures here?
boreal snow
<https://github.com/RVC-Project/Retrieval-based-Voice-Conversion-WebUI/releases/...
there is only source code?
oh nvrmind
i found it
twilit ferry
2.1.4
rare gobletBOT
Ayo? @twilit ferry level 2 !!!
karmic mantle
my voice changer sizzles on discord how can i fix it
twilit ferry
where do i find the new one?
karmic mantle
my voice changer sizzles on discord how can i fix it
i fixed it but it now echoes...
fossil vine
I know AMD is gonna be slow, but at the very least would Okada's voice changer work if i'm using AMD and install it locally?
proper shale
I know AMD is gonna be slow, but at the very least would Okada's voice changer w...
Depends on the GPU mostly
Recent AMD cards should be ok
fossil vine
Well i have an AMD ryzen 5
carmine relic
Guys some models I see are RVC V2, at 250 epochs and some at 300, what is the relation of Epochs to the quality of voice, what does that mean? how do I know what settings I'm using best for the right Epoch amount? Can someone explain, please.
Is even RVC the best voice? or is there better models?
fossil vine
Yeah i dunno how i can check the gpu, or if my device even has one
proper shale
Guys some models I see are RVC V2, at 250 epochs and some at 300, what is the re...
-
An epoch is basically how many times your dataset has been trained fully. E.g: 200 epochs means the dataset has been trained fully 200 times.
-
My recommendation is to go for 500 and check the graphs
carmine relic
I have RTX 4070 TI, and Amd ryzen 9, 5800
proper shale
There is no right amount of epochs for each model cuz every model is different
proper shale
Is even RVC the best voice? or is there better models?
It's one of the best yeah
carmine relic
There is no right amount of epochs for each model cuz every model is different
Where can I find free voice models at this level?
is there a collection or voices I can browse from?
that are english speakers? male?
fossil vine
so I suppose you have integrated graphics?
I'm assuming it might be integrated graphics, yeah
carmine relic
<#1175430844685484042>? weights.gg?
is there filter or option to set it at 500 epochs for the search?
I'm not seeing this option and or struggling to find 500 epoch sound/model
proper shale
I'm assuming it might be integrated graphics, yeah
It's gonna be rough I guess
proper shale
I'm not seeing this option and or struggling to find 500 epoch sound/model
I thought you were planning to train models. Anyways, epoch count doesn't really matter
fossil vine
I was going for wokada, but if there's alternatives that would make me sound decent and allow me to input custom models, i'm open to suggestions
carmine relic
I thought you were planning to train models. Anyways, epoch count doesn't really...
I'm going to train one to see, and also to import a model, I use the .ph at the model and then the .index at the index title when importing in RVC using w-oakda, right? I already did this but just want to confirm this, because I dunno I feel like others are having better results, where as Im doing somethiing wrong
carmine relic
@proper shale
is this response based on the connection to the server we're connect to?
or is this based on my computer specs?
if I plug my computer into ethenet will that fasten the speed/res?
proper shale
I'm going to train one to see, and also to import a model, I use the .ph at the ...
Correct
proper shale
is this response based on the connection to the server we're connect to?
or is t...
Based on your GPU's processing power and how fast it can infer
boreal snow
how many audio files do i need to train a voice model? and how long do i need them to be?
azure marshBOT
RVC Guides (How to Make AI Cover)
low shard
how many audio files do i need to train a voice model? and how long do i need th...
For help for traning a model go to #📑│making-models ,btw the dataset (audio files) dont need a specific number,it can be one with all voice lines in one or different audios,that depends how you want to put them but doesn't really matter,also the more you have (the more long) the better it is
carmine relic
guys how do I delete uplodaded models? from o-akada, it feels like I tested alot, should of named them properly, but how Do I delete them?
molten pecan
how many audio files do i need to train a voice model? and how long do i need th...
Just one audio file is enough.
For the appropiate dataset length, take this as a reference:
5m = bad model
5m to 10m = passable model
10m to 30m = decent model
30m or more = very good model
This might not always be true, the quality of the dataset plays a huge role on the final quality. A 5 minute dataset can be enough sometimes
For the quality, the vocals must be clean, free of noise, reverb, voice effects, overlapping voices, and in lossless quality if possible. Make sure the person vocalizes every vowel clearly and hit low and high notes
carmine relic
Just one audio file is enough.
@molten pecan how do I fasten the response or fast speed to output?
it seems to be delayed 1 second.
molten pecan
it seems to be delayed 1 second.
You using the realtime voice changer?
It's recommended that your PC has good specs to begin with
If you are using a Bluetooth mic, then plug it in
And make sure you are using f0 rmvpe algorithm
Idk much about the voice changer so ask on #🔍│help-w-okada
@copper pebble is this channel only about installing RVC and inference?
Or about any software related to RVC?
copper pebble
training and inference general help
molten pecan
training and inference general help
Because a lot of people come here asking about the realtime voice changer. I think we should rename the w-okada channel to #w-okada-voice-changer
Don't you think?
karmic mantle
my voice cuts out at the end of the talk its like this: my voice cuts out at the end of the tal-
restive yacht
my voice cuts out at the end of the talk its like this: my voice cuts out at the...
are you talking about about using okada
rare gobletBOT
Ayo? @restive yacht level 4 !!!
restive yacht
or when you infer models
karmic mantle
didnt get it...
restive yacht
what software are you using
karmic mantle
windows 11
restive yacht
urm to make models what program are u using
restive yacht
nah your good
restive yacht
lemme have a look rq
karmic mantle
how can i show it- (20 iq)
rancid pine
Screenshot
karmic mantle
oh
restive yacht
lol i was just looking through there github
karmic mantle
...
karmic mantle
sure wait
molten pecan
downloaded it from github
Was it called w-okada or voice changer?
If that's the case you are using the realtime voice changer. Ask on #🔍│help-w-okada for help
restive yacht
lmao i tried to start w that but it got confusing
molten pecan
karmic mantle
okayy
brittle wing
wait with this you can import your own voice models or no?
karmic mantle
wait with this you can import your own voice models or no?
i can
molten pecan
now i wish i didnt sell my pc
You can do it through Google Colab if you don't have a great PC
brittle wing
You can do it through Google Colab if you don't have a great PC
no its not that
i sold my pc because i broke my monitor from stupid game called fortnite💀
karmic mantle
i sold my pc because i broke my monitor from stupid game called fortnite💀
...
molten pecan
i sold my pc because i broke my monitor from stupid game called fortnite💀
Wait... you sold your PC because the monitor broke? You know you can just buy another monitor, right? It's not like the monitor is part of the PC
brittle wing
Wait... you sold your PC because the monitor broke? You know you can just buy an...
yeah ik
rare gobletBOT
Ayo? @brittle wing level 1 !!!
molten pecan
It's like saying you sold your PC because you broke the mouse
brittle wing
i just felt like it because i broke allat of stuff bc of fortnite💀
well i did have a 360hz monitor
molten pecan
You broke the PC too?
brittle wing
no lmfao😭
just the monitor lmfao
im saying i didnt feel like replacing once again another monitor bc of fortnite 🤣
thats why i sold my pc because i just kept on replacing monitors
soft tiger
iu still can figure out this software
odd shale
@jovial pollen
soft tiger
i tried downloading original rvc and following the steps but i dont get it
can anyone please help
odd shale
@viscid moss @molten pecan Oigan ustedes dos. Un amigo mío necesita su ayuda.
He needs help with w-okada.
molten pecan
<@274566299349155851> <@791013455251374151> Oigan ustedes dos. Un amigo mío nece...
No sé usar okada 😭 pero qué sucede?
molten pecan
i tried downloading original rvc and following the steps but i dont get it
What's the matter?
jovial pollen
He needs help with w-okada.
oh yeah #🔍│help-w-okada
molten pecan
Where are you stuck?
soft tiger
What's the matter?
can you join call so i can explain better
molten pecan
can you join call so i can explain better
You can send screenshots if you want here
odd shale
No sé usar okada 😭 pero qué sucede?
Solo quiere consejos de como hacer que le vaya más fluido.
soft tiger
You can send screenshots if you want here
i still think call is better
odd shale
-realtime
azure marshBOT
-realtime
This interaction has expired, use the command
/guides realtimeif you wish to see it again.
jovial pollen
i said everything that is needed for you to understand in help w-okada Julia
soft tiger
these ss arnt gonna make any sense
jovial pollen
if you wanna help out !
odd shale
@viscid moss
molten pecan
if you wanna help out !
Idk about okada, sorry
jovial pollen
Idk about okada, sorry
Ah no worries, you never used the realtime of RVC ?
molten pecan
Ah no worries, you never used the realtime of RVC ?
brittle wing
Sounds like the solution here was to stop playing fortnight
i didnt want to lmfao i had 98k on yt i was slowly growing but i did what i had to🤣
molten pecan
i didnt want to lmfao i had 98k on yt i was slowly growing but i did what i had ...
@half cove
slim geyser
i didnt want to lmfao i had 98k on yt i was slowly growing but i did what i had ...
Did you even make money from that with all the monitors you had to buy? 
brittle wing
Did you even make money from that with all the monitors you had to buy? <:trolle...
i made enough
supple ibex
dataset vs output, why does the output sound so bad (400 epochs, 30 minute dataset, 48k sample rate)
brittle wing
plus i have a job 🤣
half cove
can you help me
whats your problem
brittle wing
my other yt has 11.4k i posted my last vid yesterday
soft tiger
whats your problem
I need help getting a software that makes mp3 files into ai voices
brittle wing
rip
half cove
I need help getting a software that makes mp3 files into ai voices
you mean rvc?
soft tiger
my other yt has 11.4k i posted my last vid yesterday
whats your yt
rare gobletBOT
Ayo? @soft tiger level 3 !!!
brittle wing
whats your yt
i deleted my other so this acc is being ran by my editor now
my yt is blvrzz
soft tiger
you mean rvc?
can you go into call so i can explain it better
soft tiger
i deleted my other so this acc is being ran by my editor now
ah intresting
half cove
can you go into call so i can explain it better
max 2
slim geyser
Fortnite youtubers be like, I hate myself every second I'm playing this game, but atleast the kids love it!!!!
brittle wing
Fortnite youtubers be like, I hate myself every second I'm playing this game, bu...
yeah i absolutely hate that game now
only thing i miss abt fortnite is chapter 1 days
half cove
@soft tiger do go-bat
go web
its a file
for amd do dml
did you download a voice model already?
go to it
supple ibex
dataset vs output, why does the output sound so bad (400 epochs, 30 minute datas...
can someone help/explain whats happening
half cove
with ur pth file
in your rvc folder it should have a folder named
weights
no in your file explorer
no let me ss rq
yes that
put in it
you also have a index?
put it in
logs
on the model pth dropdown
choose the one u want
the infering voice
top left
do autodetect index
now ur audio
go on it
usually wav is better than mp3 but both work
molten pecan
usually wav is better than mp3 but both work
So many texts ogm 😭
half cove
copy the path of the file
half cove
So many texts ogm 😭
lmfao
soft tiger
lol
half cove
paste it into the audio path
not the folder
the file itself
copy as path
hmm
go on properties
try
H:/Ai sudios/untitled
convert
hmm
i think its cuz im windows 11
soft tiger
half cove
yes
molten pecan
i think its cuz im windows 11
@true vortex thoughts on win 11?
What would you rate it out of 10?
Should I upgrade to win 101
half cove
if u ask me any of the options i will tell u
tame birch
guys when i try to open the voice changer with the start_http thing it wont start, what to do?
half cove
yea ig
brittle wing
what tutorial should i watch
half cove
always v2
brittle wing
theres lots of different tutorials
half cove
where the model gets output
half cove
dw bout that
half cove
hey poopmaster can you show me the list again
the links?
brittle wing
yeah download links
half cove
-links
azure marshBOT
brittle wing
what i downloaded was original rvc
does it matter in tutorials
or should i find a tutorial especially for original rvc
brittle wing
are you talking to me rn
half cove
yea
half cove
are you talking to me rn
nope
half cove
an epoch is a run of the dataset
half cove
this is #✨│ai-help
soft tiger
lol
soft tiger
this is <#1159290139609137264>
real