#✨│ai-help
1 messages · Page 163 of 1
RuntimeError: PytorchStreamReader failed reading zip archive: failed finding central directory
What do I do?
im having a problem too can someone please help me im trying the real time voice changer and know i have this error im using steelsieres
changing the output or input dosnt work for me and the interned says theres some channel stuff but im a smooth brain and idk what he means by that
i also have this issue
and this
[Voice Changer][ServerDevice][audioOutput_callback] ex: could not broadcast input array from shape (1024,2) into shape (4704,2)
[Voice Changer][ServerDevice][audioOutput_callback] ex: could not broadcast input array from shape (1024,2) into shape (4704,2)
[Voice Changer] server audio performance [0, 0, 0, 0]
status: started:1, model_sr:40000, chunk:40
input : id:1, sr:44100, ch:2
output : id:6, sr:44100, ch:2
[Voice Changer][ServerDevice][audioOutput_callback] ex: could not broadcast input array from shape (1024,2) into shape (4704,2)
[Voice Changer][ServerDevice][audioOutput_callback] ex: could not broadcast input array from shape (1024,2) into shape (4704,2)
[Voice Changer][ServerDevice][audioOutput_callback] ex: could not broadcast input array from shape (1024,2) into shape (4704,2)
[Voice Changer][ServerDevice][audioOutput_callback] ex: could not broadcast input array from shape (1024,2) into shape (4704,2)
[Voice Changer][ServerDevice][audioOutput_callback] ex: could not broadcast input array from shape (1024,2) into shape (4704,2)
[Voice Changer][ServerDevice][audioOutput_callback] ex: could not broadcast input array from shape (1024,2) into shape (4704,2)
[Voice Changer][ServerDevice][audioOutput_callback] ex: could not broadcast input array from shape (1024,2) into shape (4704,2)
[Voice Changer][ServerDevice][audioOutput_callback] ex: could not broadcast input array from shape (1024,2) into shape (4704,2)
[Voice Changer][ServerDevice][audioOutput_callback] ex: could not broadcast input array from shape (1024,2) into shape (4704,2)
[Voice Changer][ServerDevice][audioOutput_callback] ex: could not broadcast input array from shape (1024,2) into shape (4704,2)
[Voice Changer][ServerDevice][audioOutput_callback] ex: could not broadcast input array from shape (1024,2) into shape (4704,2)
don't use server audio, use client
same thing, just use client
what do you mean client or server?
this happened because you combined 2 diff types of input/output
there's two types of output basically, client & server
in the gui
server has a few benefits but client is just better overall
i ended up just deleting whole folder
Ayo? @mossy sun level 1 !!! 
yeah hahahaha
anyways use the same type for both, in the end they have the (mme) type for example, so you'll select only mme output/inputs
and im trying for hours now how i can get this running so people can hear it
or whatever you want to use
oh ok nice thanks
you're welcome
You should delete pretrain folder, start the app and wait until it downloads the files
what is a good local TTS option ?
how do i make it so people can hear me with the model on ?
you need a virtual audio cable, check the guide (has download links and it's pretty easy to set up)
EOL - No further Updates
Github - Blanc-dot
Discord User ID - 824922747423031359
Despite being end of life, most if not all information has not really changed, so should be very accurate until actual new stuff comes out.
Other Links
Antasma's Local Error Fixes
Antasma's Colab guide
Sushi's useful...
why my gpu is not detected ? btw my gpu is amd rx 6500 xt , anyone can help ?
Does the index help me to find the original voice
What's the problem?
Index controls the accent
Mmmm
Are you arabian
No
Ok
But how do I get a child's voice on RVC
I mean how i get child voice in RVC
And this happens when I try to do the same thing, but with huggingface.
Ayo? @night rune level 2 !!! 
I'm trying to download models made by myself.
Nahhhhhhhhh you didn't
Xdddddd
youre disgusting
No
Because i tried to get original voice in RVC
dose anyone know why when i try using a voice most of them sound nothing like audio clips i herd. can i fix this
is there a way to run the realtime rvc gui without the command prompt window?
EOL - No further Updates
Github - Blanc-dot
Discord User ID - 824922747423031359
Despite being end of life, most if not all information has not really changed, so should be very accurate until actual new stuff comes out.
Other Links
Antasma's Local Error Fixes
Antasma's Colab guide
Sushi's useful...
ah ok thanks
what is the best webui or software to make ai covers using existing rvc models, (im doing it locally) because i don't want to rely on weights.gg for everything
ok I think I'm doing it wrong, when I try to run the VBS script it seems to do nothing, but running the original .bat it's pointing to still works fine
This is my VBS script:
Set WshShell = CreateObject("WScript.Shell")
WshShell.Run chr(34) & "C:\AI stuf\RVC1006Nvidia\go-realtime-gui.bat" & Chr(34), 0
Set WshShell = Nothing
Ayo? @nocturne talon level 1 !!! 
-local
Suggestions for @mental wagon
- 🍏 Applio, by IA Hispano GitHub
- Mangio-RVC-Fork, by Mangio621 Huggingface
- RVC Studio, by SayanoAI Huggingface
- AICoverGen, by SociallyIneptWeeb GitHub
- Replay, by Replay Team Website
- Original RVC, by the RVC-Project team GitHub
- GPT-SoVITS, by RVC-Boss GitHub
Credits to Faze Masta and Antasma for compiling these links.
Can’t go wrong with Original RVC
In the guide it says to remove reverb before extracting main vocals. But what if I'm after backing vocals? A lot of backing vocals can be more reverbey and echoey than the leads. Would dereverbing first end up messing up the bgvs?
should be fine i guess
actually that's one part of the guide that needs to be revised, I'd suggest removing reverb AFTER extracting leads and backing vocals
Oof yea i just did some dereverb/echo and it took off so much BGVs. It's almost like a weird karaoke lite. But wait why wouldn't you dereverb first if you're aiming for just the lead vocals though? The dereverb isn't taking out any lead vocals?
Ayo? @glossy fjord level 1 !!! 
i do it after to not lose any quality after i separate leads/bgvs
but honestly i focus more on the leads in this case
-kaggle @gusty scroll
Suggestions for @gusty scroll
- 🆕 UVR5 NO UI by Eddy
- How to use RVC Mainline Kaggle by Cauthess
- 🍏 Applio RVC by IA Hispano
- ✨ RVC Mainline by Hina
I havent used colab in a while and now it doesn't work even though I have pro. Does colab not work anymore in general? For some reason I thought it was just banned for the free tier.
what colab u used?
!colab
All I'm looking to do is make inferences based on existing RVC models.
Table Of Contents Introduction Model Loader (Download & Upload) Inference (use RVC AI Voice Models) Ilaria TTS Settings (Inference) Introduction Ilaria RVC Zero, is an RVC (Retrieval-based Voice Conversion) Fork made by Ilaria & mikus, running only on Hugging Face Spaces, it’s called this way...
Hey guys, does anyone know why my custom models don't work but the default ones do?
how do install pytorch it will not let me
why does this actually sound like its not ai
Ayo? @agile apex level 2 !!! 
Thank you - the Ilaria RVC Zero is incredible. So fast and way easier than Colab.
How do I not make the AI model I trained not sound like a fucking static beep?
Is it possible to change the "32000" to "48000"
I really want to merge together 2 models, and I need to have both be on "48000" or both be on "32000" for me to merge them.
thank you 💖
Just asking out of curiosity. Is there any way to use a AI voice for a TTS mascot on a stream? I seen ones like Tangia but for Tangia they don't let you use models and ask you to read a script so I wonder if there's any program that actually lets you use voice models.
For best, stable and smooth results
Should you use rvc with ur gpu at 50-90% utilisation
Or below 50% utilisation
Or it doesn't matter at all
I need help
- Applio, by IA Hispano Google Colab
- RVC Disconnected, by Kit Lemonfoot Google Colab
- RVC Mainline, by Hina Google Colab
- AICoverGen-WebUI, modded by Hina Google Colab
- AICoverGen-NoWebUI [English], by Ardha, fixed by Eddy, Hina and Gdr Google Colab
- AICoverGen-NoWebUI [Spanish], credits to Eddy, Hina and Gdr for translating and fixing Google Colab
- Ilaria RVC, by thestingerx Google Colab
- EasyGUI, by rejects Google Colab
- UVR5 NO UI for Google Colab, by Eddy Google Colab
- 🆕 UVR5 UI, by Eddy Google Colab
so basically, realtime tts with rvc voices on stream?
Pretty much yeah. I doubt its feasable but curious anyway.
Ayo? @chilly pebble level 1 !!! 
It could be possible, i dont think so with wokada (realtime voice changer with rvc but no ttts), but you can just use forks such as Ilaria RVC or Applio for tts with rvc and then follow this to use tts on calls
Table Of Contents Introduction Index of the best TTS 1. ElevenLabs/11Labs: 2. Bark TTS:Bark is a multilingual TTS model created by Suno AI. It’s good emotionally, There are Non-Speech Sounds you can use: 3. Edge TTS: 4. StyleTTS2: 6. XTTS2: 8. MetaVoice: 9. MeloTTS:MeloTTS is a high-quality mu...
also reminder that its pc only, if its a good one you can use ilaria rvc or applio locally, else do it online
Btw it can't really be tts realtime as you have to write ofc 
@low shard Oh I meant as in have a redeem for AI TTS sorry.
what do you mean with 'redeem' ?
So viewers can redeem TTS/use it to read text but with RVC. Sorry. Lemme try to get a example if I can't explain it. 🤔
so basically, have a live where whatever viewers type in the chat, and whatever they write gets transformed into a tts rvc voice?
Lemme try to get an example sorry.
its fine
Whoever knows the creator of this Colab should tell him that cuda libraries have problems https://colab.research.google.com/drive/1qezaBvCn4eSG1U2L4a64tEHTynbIu8eN
@viscid moss check out your colab
its working but doesn't use cuda gpu, only cpu
the errors seems to be:
Looking in indexes: https://aiinfra.pkgs.visualstudio.com/PublicPackages/_packaging/ort-cuda-12-nightly/pypi/simple/
ERROR: Could not find a version that satisfies the requirement ort-nightly-gpu (from versions: none)
ERROR: No matching distribution found for ort-nightly-gpu
related to:
!python -m pip install ort-nightly-gpu --index-url=https://aiinfra.pkgs.visualstudio.com/PublicPackages/_packaging/ort-cuda-12-nightly/pypi/simple/
checking..
If the problem is solved, I would like you to ping me.
sure, I'm working on the fix rn
It's ok as long as you don't hit 100% utilization
If you're talking about realtime
i think i found a fix
I just move the installation to other cell and it seems to be working
what was ur fix?
replaced that line that causes the error with !sudo apt install -y nvidia-cuda-toolkit
installation time in..?
did you just do it in the other cells? or modify smt of that line too
just move
when that line is in the same cell, the indexes cannot be found
idk why, this notebook has been working for months with the same code lmao
now is working again :3
Solved! Try the notebook again
I'm on it
is it faster or slower than your fix?
same but it's better 'cause same cell
Ohh, btw shouldn't the original aicovergen (english one) have that issue too?
i never really used it so idk if its different or same code
Additional requirements are necessary?
yep same issue also karafan

yep
I'm editing the notebook right now to make it back to a single cell
are you the owner of those other 2 colabs too?
yep
CoverGen en and es problem solved, just testing to make sure
i'll edit karafan in a bit
As soon as the Additional Requirements part starts, it gives an error
Maybe it's because I'm editing the notebook rn
Alright
Ok, I'll just wait until it's ready so I don't use up my GPU time
I will appreciate it
Done! Make sure the colab looks like this:
that's how it is
Ayo? @crimson depot level 2 !!! 
lmk if solved
Installing dependencies, I'll let you know
It is working as usual. Nice work!
-colab
- Applio, by IA Hispano Google Colab
- RVC Disconnected, by Kit Lemonfoot Google Colab
- RVC Mainline, by Hina Google Colab
- AICoverGen-WebUI, modded by Hina Google Colab
- AICoverGen-NoWebUI [English], by Ardha, fixed by Eddy, Hina and Gdr Google Colab
- AICoverGen-NoWebUI [Spanish], credits to Eddy, Hina and Gdr for translating and fixing Google Colab
- Ilaria RVC, by thestingerx Google Colab
- EasyGUI, by rejects Google Colab
- UVR5 NO UI for Google Colab, by Eddy Google Colab
- 🆕 UVR5 UI, by Eddy Google Colab
Ayo? @vale osprey level 2 !!! 
Bro, I downloaded the voice of the girl, but here is only the path, there is also an index needed
you don't necessarily need index
just upload the pth and it should work
do i need to redo feature extraction when i want to continue training a model?
no, just input the same configurations as last time and click train. if the checkpoints are present it'll resume
thanks 👍
I still need help
Beginner: https://docs.aihub.wtf/essentials/how-to-make-voice-models/
Advanced: https://rentry.co/RVC-dataset-RX11
We can't force you to read through it, but this everything you need to know about good model making
volume input on my mic shows as negative numbers, which means the voice changer cannot find my voice. tried using clownfish to enhance but its still not enough
if anyone has a fix
would be appreciated
for make song with rvc what is the best ? harvest dio pm crepe crepe-tiny ?
Hello! Curious on what apps could use the RVC models, currently im trying with voice.ai but does someone have a good recommendation for a voice changer apart from this one?
I'm trying to use the "AICoverGen WebUI" Colab (since I'm running AMD), and it keeps breaking with this:
Timer: 00:33:022024-07-19 18:31:52.375858025 [E:onnxruntime:Default, provider_bridge_ort.cc:1745 TryGetProviderInfo_CUDA] /onnxruntime_src/onnxruntime/core/session/provider_bridge_ort.cc:1426 onnxruntime::Provider& onnxruntime::ProviderLibrary::Get() [ONNXRuntimeError] : 1 : FAIL : Failed to load library libonnxruntime_providers_cuda.so with error: libcufft.so.10: cannot open shared object file: No such file or directory
2024-07-19 18:31:52.375898490 [W:onnxruntime:Default, onnxruntime_pybind_state.cc:895 CreateExecutionProviderInstance] Failed to create CUDAExecutionProvider. Please reference https://onnxruntime.ai/docs/execution-providers/CUDA-ExecutionProvider.html#requirementsto ensure all dependencies are met.
```How do I make it work, or is there a newer version out?
Use Ilaria RVC on hf instead to generate covers
RVC Realtime and W-okadas
Where can I find it?
-spaces
Suggestions for @grizzled cove
- 🆕 UVR5 UI, by Eddy and Ilaria Huggingface Spaces
- Ilaria RVC Zero, by Ilaria Huggingface Spaces
- RVC⚡ZERO, by r3gm Huggingface Spaces
- Applio, by IA Hispano Huggingface Spaces
- AICoverGen, by r3gm Huggingface Spaces
- Advanced RVC Inference, by r3gm Huggingface Spaces
- RVC v2 Huggingface version, by Clebersla Huggingface Spaces
Thank you!
What settings do I have to use? I uploaded a model, a song, and tried it with the Pitch level set to "1" (because male to female voice conversion) and it sounds terrible.
Ayo? @grizzled cove level 1 !!! 
It seems like it doesn't even separate the voice and music, but just covers both?
I don't get it
You upload vocals only
You can use UVR, Mvsep, or xminus to separate music and vocals
Ah okay. I thought it did it the same way the one I used before did, which separated voice and music, then converts the voice, and puts it together immediately. Thanks!
thanks! Ill have a look
is there a feminine voice that doesn't break with laugh?
Hi guys, I'm trying to do inference on a hugging face space. But the index file of the Usopp model I found doesn't work: https://huggingface.co/FrappeChan/Usopp_English
Even if I try to upload it manually, there's an error. Any chance this could be a simple fix?
Is the index file start with added or something else
I would suggest you to use Windows 10
Actually, you're using a wrong link.
The model link must be like this: https://huggingface.co/FrappeChan/Usopp_English/resolve/main/Usopp_English.zip
Nope, laughing normally isn't possible on W-Okada.
ok im switching to windows 10
cuz windows 11 is so garbage
It was just my opinion but go ahead.
I agree that W11 is trash
I'm not sure.
Linux FTW
ok
i swear i broke my pc for linux
and i redownloaded it
it worked
but linux does not have anticheat games so i returned to windows
i play league of legends and valorant
so ye..
I agree that Microsoft's handling of Windows 11 has gotten worse and worse.
When I need Windows, I pop a virtual machine instead. It's basically a fake computer.
virtual box?
Yeah, VirtualBox, VMWare, all that goodie.
Some you can do GPU passthrough, where you attach it to the guest machine. VirtualBox isn't one of them sadly. Be aware you can't use the card on the host.
i dont have powerfull device anyways
so ye
i tried virtual box
its good but not like the host ofc
Though giving the guest access to the host GPU can bring native-like performance.
I switched to Linux as my daily driver as Windows 7 end-of-life approached. No regrets whatsoever. Definitely give it a try at least once sometime in the future.
Gaming is often considered the biggest barrier to switching; I don't blame them.
gaming on linux is good
i think they use proton or smth
u can play steam games native
riot games u cant sadly
cuz vunguard anticheat
It's gotten better, but I can agree some developers don't like it and catch players for cheating when there's no cheating.
rare
riot vunguard is trash ngl but it is what it is
im froced to use it
The index file itself is called "model.index", similar to the pth file "model.pth".
I appreciate everything you do for us hobbyists. Any advice on how to get fusion of two models working in Ilaria ? Seems like no matter which models I choose the architectures are not compatible. I'm not sure what I should be looking for to know if they'll match. I tried a buncha different combinations and none worked.
No I just added this link here for yall to see the card on hugging face, I'm using the right link when putting it into rvc
needs to be .pth files of the same model since you're merging the best G and mel/kl
its currently broken, will be fixed when possible :3
Understood. Out of curiosity how well does fusion of two models work? Does it come out sounding less human for singing? Do they need to be somewhat similar voices?
it depends, its kinda a mistery, you have to try it
Anything I can do to make the voice sound less roboty? Every song sounds robot-like. Especially "sh" and "s" sounds.
bro's late
no clue i don't use it
ryzen 5 5600g with radeon graphic card
oh k
@wispy lodge
where i can download real voice changer?
should be good enough with this fork https://github.com/deiteris/voice-changer
what i get?
amd64-dml
it's a zip file so you need to extract somewhere (either of root drive C:, D:, etc recommended)
Ayo? @knotty moth level 18 !!! 
so its does not matter
ok
thx
@knotty moth
is it normal?
it downloads a bit
Limit frame rate or settings on your game
Is there a good RVC fusion model maker hosted on huggingface?
tried using the fumiama fork of rvc (recommended in the rvc discord) and i seem unable to train using an amd gpu (linux, rocm). anyone know a fix?
Which AMD GPU? Do you have the version of PyTorch with ROCm acceleration?
yep, made sure. 6700 xt. worked before with the original RVC-Project repo
I have 2 5500 XTs; Fumiama's RVC training works for me.
Ayo? @wide ingot level 20 !!! 
which version of python did you use?
i had dependency issues with 3.8 but none on 3.10, but i'm aware 3.8 is recommended
Python 3.10.13 + PyTorch 2.0.0 + ROCm 5.2
Grab the one that has "rocm" in it, or you won't able to train.
yeah ofc, always just been grabbing it from pytorch.org
Also, use a known ROCm version that works with your card.
Since you have an RX 6000 series card, you may be aware of using an environment variable to override the architecture version.
I had to do that as well.
yeah i did that ages ago
Ayo? @little drum level 3 !!! 
i've trained successfully before on the same gpu and pc
But now, I have built the ROCm components completely from source to target my card, and no more of that environment variable shenanigans. It just works.
yeah torch seems to be the key here as i went back to the rvc repo i used before and shit's fucked
thanks for the advice
It's not for the faint-hearted, but at least I can use the latest PyTorch.
Sorry, @crimson wraith, didn't mean to reply. I don't know the answer to your question.
wasn't working on rvc-project but fumiama seems to be picking up on my gpu. thank you so much for that
I'm thrilled to announce I'll r@p3 all of the staff
Better not tag me in announcements again bitch ass bitches
?
It kinda depends on the settings. Though with igpu it may be difficult to achieve
Hello Can anyone help me? I want to buy a new computer and am thinking of choosing an AMD or Intel CPU Which one is better?
wait for Ryzen 9000-series (Zen 5) launch in the next 2 weeks
My budget so low 😭
Using the new audiofish rvc model creator.
Can i export the model as a pth and index file?
And try the channel below it
Who is "child" anyway?
Ayo? @ebon tulip level 1 !!! 
Not a specific character?
Ayo? @forest girder level 1 !!! 
Try https://voice-models.com/
Its a good website
the thing is
there's no example in this website
you'd have to download and try your luck
& the names are weird
You'll have to try.
Using the new audiofish rvc model creator.
Can i export the model as a pth and index file?
It's a quality measure
200 is good
More is usually better
No. But before you download check the epochs
alright, but each model has two files, index and pth, which one shall I use?
What app are you using
weights.gg is another alternative
Hey... What the heck are you doing with a 10 year old voice and a voice changer???!!!
Ayo? @ebon tulip level 2 !!! 
Anyway. You should throw the whole folder in the models folder
Un compressed
Ayo? @forest girder level 2 !!! 
And... How do i download okada??
define your budget range, if you want to keep your motherboard, you'd rather stick on the old Intel 12-th gen or AMD 5000-series, otherwise grab an AM5 mobo upon some nice price drop.
you can grab 7600X but if you want to futureproof whenever getting more budget, yeah wait for the zen 5 cpu instead.
I have a question, is there a way to make this run in realtime? (like it would change your voice in realtime)
- Applio, by IA Hispano Google Colab
- RVC Disconnected, by Kit Lemonfoot Google Colab
- RVC Mainline, by Hina Google Colab
- AICoverGen-WebUI, modded by Hina Google Colab
- AICoverGen-NoWebUI [English], by Ardha, fixed by Eddy, Hina and Gdr Google Colab
- AICoverGen-NoWebUI [Spanish], credits to Eddy, Hina and Gdr for translating and fixing Google Colab
- Ilaria RVC, by thestingerx Google Colab
- EasyGUI, by rejects Google Colab
- UVR5 NO UI for Google Colab, by Eddy Google Colab
- 🆕 UVR5 UI, by Eddy Google Colab
I think that'll help, thanks.
hello, can someone help me ?
yup
-realtime
This interaction has expired, use the command
/guides realtimeif you wish to see it again.
Hey, i've just trained a model, but when i put it into Replay, it says this: No valid .pth files
Am i missing something?
not sure.. never really used replay myself
Cant send screenshots else i'd ask whenever i'm actually using the proper file lol
the model file that is used is located on a folder called weights
most of the time
never in logs
thats the files needed to resume if you want to continue training the model
aight have fun
thank you so much but i just found out how to fix my problem
📚 All-In-One English documentation
Full AI Voice Model Training Guide (Local)
Link: YouTube
Credits: Christopher Villanueva
Model training with Mainline RVC
Link: Rentry
credits: Raven (ravencutie21)
AICoverGen Colab Guide
Link: Google Docs
Credits: Eddy (Spanish Helper)
Create a model with RVC disconnected (colab)
Link: Google Docs
Credits: Angetyde
best Search feature ratio settings?
For real time just set it to 0 but for inferencing you could set it to like 0.3 or 0.5
What would be the best way to separate intrumental audio from the voices in a song?
check websites like minus x
this would help too
I remember I used an app to do it since rvc wasn't really that good and the app was perfect, but I just can't remember what it was
these websites will add a few more steps but they are easier
I see, thanks
just to be clear minus x does have a limit of 10 minutes
this one dosent
Is minux x better for shorter videos?
no its basicaly worse in every aspect
i see, so always use vocal remove
bot not working?
yeah. and for re combinig vocals and music use audacity
its a free windows app
i always try to help as much as i can
Will there be any problems if I install rvc realtime in an external 1 tb hdd with a max write speed of 130 mbps
Hey everyone, I just made another model which has 750 epochs. The problem is, I cant find its index file. Where would Applio store it, and does it even make an index file by default?
should be in the logs folder
if not you forgot to train it maybe
might be slower to write those d/g files
but it should be fine for the most part
I have the actual pth model, but I cant find its corresponding index. I looked in the logs folder already too
well, unless you think about uploading the model to here, the index is not really necessary
Is an index something I also have to train specifically?
yeah on the train tab it should have the "train index" button
I was kinda thinking about uploading it here, hence the question. But if not its really no sweat
ah
unless they removed it for god knows what reason 😭
How to fix the voice crack when I create AI cover?
Oh, is that the "Train Feature Index"? I was wondering what difference that was
could be background vocals, make sure you're using lead vocals only
yes, thats it
dang, okay thank you so much!
It's like a high pitch voice crack
then mess with the pitch
I should be able to train that faster off of a pretrained RVC model right? Like if I tell it to restart from a checkpoint or something?
you're welcome :)
you don't really need to select pretraineds for it
Are the indexes faster to train?
just keep the same settings and model name and it should do it just fine 
yeah
like
10 seconds 💀
I don't know what do do. Should I show you the sample of my problem?
yeah, send on dms
Well I really appreciate your help!
:) always glad to help
Literally already done 💀
Ayo? @slate birch level 3 !!! 
What is it that Indexes do exactly? I kinda beat myself up about them and assume their lack of use is a cause for artifacting or cracking or things like that. Is that actually the case?
index helps with the model's pronunciation, applies the accent in the final result
how do i know if im overtraining or over training them?
check the tensorboard graphs
most of the time you aren't overtraining though, if you're on like 200-300 epochs
but ofc each case is different
im looking at the graphs but i dont understand them
so yeah check the tensorboard graphs, g/mel and g/kl mainly, if these go up for an extended period of time, yes it's probably screwing up
Is overtraining a bad thing? Or is it just about eliminating redundancy
this is okay right?
Where do I put the .index file so it is detected by the web gui?
i think weights(not sure)
no, that's where I put the pth files
let me check
that's gradients
overtraining IS bad, yes
i dont rember
it's just hard to get to that point
I see, though, how much does the index file help?
How so?
because if you overtrain, it'll be quite good with its own dataset, but with any new audio, it just screws up hard
I found out where to put it, I need to go into the logs folder, that's where i put the .index files
Ayo? @bleak dust level 2 !!! 
yeah
different pros and cons. in general RMVPE is far easier to use than crepe as it's more forgiving towards audio that has imperfections. in most cases you're likely better off just using RMVPE because of that
Alright, thanks
when i upload audio to Ilaria it doesnt load
nvm just converted the wav to mp3
oh ok
when i tried to do a song with covergen, the files werent there after finishing, im trying to do it again but after less than a second it gives me
python3: can't open file '/content/src/main.py': [Errno 2] No such file or directory
2
how do i fix that?
Guys, is there anyone who can help me? When I press convert in the RVC GUI, it seems to be working for a while, but then it gives an error. It gives this error.
RuntimeError: CUDA error: the launch timed out and was terminated
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.
What should I do?
-gui
it's outdated, read the guide for up to date ones https://docs.aihub.wtf/
Last update: Mar 10, 2024
Guys I’m curious does RVC support all languages? Are the English models higher quality than other ones?
Rvc gui is outdated
Should I download a new one?
Ayo? @wooden reef level 1 !!! 
you need training?
Do you need training or inference
inference
easy cloud solution https://docs.google.com/document/d/1YbXcLFPaGjhOdG5NFkK3QrucCEpHZBwFUxkeMO8aB18
Table Of Contents Introduction Model Loader (Download & Upload) Inference (use RVC AI Voice Models) Ilaria TTS Settings (Inference) Introduction Ilaria RVC Zero, is an RVC (Retrieval-based Voice Conversion) Fork made by Ilaria & mikus, running only on Hugging Face Spaces, it’s called this way...
-spaces I recommend Ilaria RVC if you only need inference since it’s online
Suggestions for @wooden reef
- 🆕 UVR5 UI, by Eddy and Ilaria Huggingface Spaces
- Ilaria RVC Zero, by Ilaria Huggingface Spaces
- RVC⚡ZERO, by r3gm Huggingface Spaces
- Applio, by IA Hispano Huggingface Spaces
- AICoverGen, by r3gm Huggingface Spaces
- Advanced RVC Inference, by r3gm Huggingface Spaces
- RVC v2 Huggingface version, by Clebersla Huggingface Spaces
thanks
-colab
- Applio, by IA Hispano Google Colab
- RVC Disconnected, by Kit Lemonfoot Google Colab
- RVC Mainline, by Hina Google Colab
- AICoverGen-WebUI, modded by Hina Google Colab
- AICoverGen-NoWebUI [English], by Ardha, fixed by Eddy, Hina and Gdr Google Colab
- AICoverGen-NoWebUI [Spanish], credits to Eddy, Hina and Gdr for translating and fixing Google Colab
- Ilaria RVC, by thestingerx Google Colab
- EasyGUI, by rejects Google Colab
- UVR5 NO UI for Google Colab, by Eddy Google Colab
- 🆕 UVR5 UI, by Eddy Google Colab
150
Last update: Feb 10, 2024
Train it till it starts overtraining, make sure to check Tensorboard.
hi
Do I have to fill this In if I'm training using my CPU
TRAINING USING YOUR CPU???
The GPU I have Isn't supported
at this point train on colab
AMD Radeon RX 6600
yea nah
but a cpu cant do it
i mean it can but it will be so slow it will kill your cpu
-colab
- Applio, by IA Hispano Google Colab
- RVC Disconnected, by Kit Lemonfoot Google Colab
- RVC Mainline, by Hina Google Colab
- AICoverGen-WebUI, modded by Hina Google Colab
- AICoverGen-NoWebUI [English], by Ardha, fixed by Eddy, Hina and Gdr Google Colab
- AICoverGen-NoWebUI [Spanish], credits to Eddy, Hina and Gdr for translating and fixing Google Colab
- Ilaria RVC, by thestingerx Google Colab
- EasyGUI, by rejects Google Colab
- UVR5 NO UI for Google Colab, by Eddy Google Colab
- 🆕 UVR5 UI, by Eddy Google Colab
rvc disconnected
Ive downloaded some models. How can I quickly telll if it is RVC v1 or v2 and if it is RVMPE or not? The model info checker I found tells me 40khz, how many epochs, and if it has pitch tracking but not which architecture it uses.
99% of the models made recently are v2, rmvpe. you don't really need to worry much about that tbf
it should work
The one I like the most is RVCv1 (David Bazan) so i was looking for other models that are RVCv1 to do a merge with to create something "unique".
But yeah seems like most of the rest are RVCv2 which makes sense.
Does anyone here sell the service of fusing models and doing a decent job of retaining quality (minimum added raspiness or robotic noises) for singing voices?
so i wanna ask why the app all voices lag and cut and the app give me error
any ideas?
you turned on every setting under the sun 
Ayo? @edgy oak level 10 !!! 
i have low end pc idk if this is good
so i wanna ask why the app all voices lag and cut and the app give me error
Ayo? @barren acorn level 5 !!! 
I only use colab so I can't answer that #🔍│help-w-okada
no
only use sup2 and echo, disable sup1
vb cable is bad, use vac lite
index 1 is excessive, index files in realtime usually destroys the model quality (not always, it depends), so use index 0.5 or don't use it at all, also using index files causes huge cpu usage on windows, tho in modded wokada you're able to run index files nicely if you have any of the latest expensive intel cpus
chunk is too high
extra is too low
They want to play games on igpu, so chunk at 512ms and extra at 1-1.5s is preferable
But index yeah, don't use it
makes sense
@chrome ember Preferably look in #1175430844685484042 and review the voice samples for yourself since everybody trains their model differently. Keep in mind not everyone has the intention of making hq models because we can't force them to improve
i have ryzen 5 5600g
apu
so i disable sup 1?
ok
yep
Extra is audio history that helps to improve quality
More - better, but higher GPU usage
how much i put it?
.
when i talk i get drop fps
even if i limit my fps to 60
You have to upload it and select it
¯_(ツ)_/¯
You can try increasing chunk more
This will decrease GPU usage by voice changer
whats the most up to date voice changer currently
Help, How can I upload a model in discord
For what?
I want to share a model I made
#1159514067187277865 /submit
okada
Not RVC realtime?
i prefer okada
Gotchu
How do i get less voice cracks like sometimes the voice is jus getting unclear and it sounds like a robot
the chunk is way too high
from 96 to 192 you should be good
also the index sometimes cause problems
guys where can i find models?
Nah i have it low, accident
ty
could also be the model not being perfect
You can search rvc ai voice model:
- #1175430844685484042
- #🔍│find-models
- https://weights.gg/
- https://huggingface.co/models (but watch out cus in hugging face there arent only rvc ai voice models)
- https://applio.org/models
- https://voice-models.com/
if there isnt one, you can:
- #1159289738314919936
- #1191429836321849435
- make it yourself with our docs guides https://docs.aihub.wtf/essentials/how-to-make-voice-models/
aight tysm
yw
I just finished training but the INDEX file isn't here
I checked every single folder to make sure It's not gone but I don't see It anywhere
The index file starts with added
Make sure you trained feature index ratio
Or train index
Oh ye I forgot about that ig
Thanks
Are voice models safe
Yes
Do i copy link or download
RVC is open source, you can view the source code
Copy link if your using ilaira RVC
Ok
And plus your using it online in a HuggingFace space
what do i do when discord doesnt pick up the cable output audio
Discord warns you at any direct download link
The “download model” section on Ilaria RVC
Where
can anyone help w why the virtual cable wont work on discord
Make sure the output on w-Okada is the cable and on discord the input is also the audio cable
Upload model?
Model downloader
Model url
The model uploader is if you have the files for the model
okay thanks!
Ayo? @elder basalt level 1 !!! 
What is the models name, input it
lego joker
Wait wait till it says model downloaded
Yes
Now go to inference
Where
How To Troubleshoot 
- Don't simply mention your issue, like "
my rvc is not working". - Describe the step you are on, what you're trying to do, the RVC you're using, a screenshot, etc.
- The more context, the better.
- Don't be desperate. You can ping a Helper, but if they ignore, they aren't available/don't know the answer.
- It's okay if you're frustrated, but don't take it into this server.
- Don't DM without prior consent.
- Don't ask for every little instruction. Put your own effort & test things by yourself.
- Don't ask to ask.
- Check if your answer is a Google search away/on our guides website.
read the part about “BE PRODUCTIVE”
Ok i will try one step on my own
I have a question
How do i open the google doc without closing the google chrome because the google docs takes over the google chrome.
The doc is just a guide for it
You can copy the link and put it in a separate browser tab
Ok
Hey boys it’s been a few months since I’ve last checked here… is the google colab version of RVC fixed?
Yup rvc disconnected is still working
-colab
- Applio, by IA Hispano Google Colab
- RVC Disconnected, by Kit Lemonfoot Google Colab
- RVC Mainline, by Hina Google Colab
- AICoverGen-WebUI, modded by Hina Google Colab
- AICoverGen-NoWebUI [English], by Ardha, fixed by Eddy, Hina and Gdr Google Colab
- AICoverGen-NoWebUI [Spanish], credits to Eddy, Hina and Gdr for translating and fixing Google Colab
- Ilaria RVC, by thestingerx Google Colab
- EasyGUI, by rejects Google Colab
- UVR5 NO UI for Google Colab, by Eddy Google Colab
- 🆕 UVR5 UI, by Eddy Google Colab
-kaggle
Suggestions for @pastel tendon
- 🆕 UVR5 NO UI by Eddy
- How to use RVC Mainline Kaggle by Cauthess
- 🍏 Applio RVC by IA Hispano
- ✨ RVC Mainline by Hina
There's also kaggle but no one has figured it out so far 
Yeah I want to figure it out as well... but sadly
this
Here's the error that I get every time
Since it's too long to send here I just sent it as an attachment
Might be because your "internet" toggle isn't turned on. You need to verify your phone number
try the mainline guide I wrote above too. I provided a lot of images to get it running
There, I updated it again so no one can get lost because they're scared of clicking cells
Never heard of Kaggle
It's one of googles hidden services for IT people "in the know". Colab is a scam
jk its for data science so you get a lot of gpu
Wow so even my scrap 2GB Nvidia GPU will work better that way I guess
It's really good. I went to sleep for that whole session and it's still training. But yeah the guide says the maximum is 12 hrs to be on the safe side
My internet toggle is on
I'm not sure. I know applio has a few bugs rn and it seems that notebook hasn't been updated in 10 months
If u want to use applio on the cloud, use colab
I already have colab but I want to use kaggle as I don't have that much money (and my colab credits are running out). Thanks for your assistance
you dont need to waste money on the colab compute units, if you wish you can switch to another google account, but yeah kaggle is a good alternative
i have shitty connection does this works?
try it
GUIs will have auto backups to your training so it's highly recommend over rvc disconnected
your internet speed doesn't matter there
only matters if you're training models (because u have to download the model xd)
really?
i used the collab 5 months ago
is it better?
not really, has a daily limit and colab pro is expensive
kaggle has better speed and longer daily limit
for free
but is way harder to setup
I re-updated updated the guide again out of someone elses request just now. You cannot get lost https://rentry.co/RVC-Mainline-Kaggle
This guide for Mainline Kaggle is an alternative option to the Mainline Colab notebook.
It is currently complete and should walk you through every step of the way since Kaggle has a difficult learning curve. It will be updated constantly since new updates to the notebook will slightly diverge fro...
i highly advise you to use kaggle's mainline notebook
believe me, you'll like mainline more than applio hehe
I mean kaggle would be better for performance
Could u try to run applio kaggle ? If it doesn't work I could tell Vidal about it
sure im currently training with mainline rn. But it's probably outdated if they already have their internet switch flipped on
you cant view the tensor board with applio kaggle, but training works just fine
you also have to use applios built in sync graphs thing bec you cant access the files like you can with mainline
how we can make an Ai cover ?
Welp I already did it like a few weeks ago (and plus I don't use it for just voice training)
Sadly google is smart and detects if i'm using an alt
Thanks for your input though!
Ayo? @manic plaza level 5 !!! 
Anyways if you use more GPUs while training, will it actually be faster
I did that before and it seemed like it was faster at first, but then I realized the speed difference wasn't that much
trying to use Ilaria RVC but it constantly gives me errors
Ilaria fusion is back up by the way
I'm probably doing something wrong, but i can't get fusion to work. Still complains about architecture mismatch even when both models are v2, 40k, and have pitch tracking.
it needs to be the same model pth probably. Thats where the compatibility comes in
Where can I read about what that means - using the same model pth? Isn't the model pth the actual sauce that makes the voice sound like it do? Im guessing you dont mean litereally the same pth file?
So say you have a g/total that stops going down on the tensorboard, save the lowest point as your epoch
now monitor your kl/mel and choose the best/lowest point as your 2nd pth file
you can merge g and mel together as a 50/50 split
just get used to creating models first before you worry about this quality of life feature ^^
and knowing how to read tensorboards
Oh my bad. You guys have sent that image over before and honestly my eyes glazed over when I saw it because I am very unfamiliar with running the tools locally. I was thinking the Ilaria fusion GUi would kinda do the tensorboard part for me. It sounds like I am going to have to step up and learn how to do this, likely using a windows machine and follow these guides. Are the things in the graphic going to be pretty intuitive to sort out once I get it running on a local machine?
yeah it's really easy to setup and we have guides here https://docs.aihub.wtf/rvc/resources/epochs--tensorboard/#tensorboard
Last update: Feb 10, 2024
and I don't need a graphics card?
for local training, you need a graphics card. If you don't have one, then use colab or kaggle
On average, how long does it take to create a fusion?
5 seconds. What you're trying to do is merge 2 different models together which probably doesn't even work
Ayo? @thin stump level 33 !!! 
Merging two different models (voices) together doesn't work? Because they're trained differently, or just in general?
I haven't tried it before but you got answer because it repeatedly told you the architecture weren't compatible
but if I do the tensorflow things you described and create a new pth file, then compatibility can be attained? My apologies for being a big dummy.
yes. Again best g/total and best mel
keep track of kl since he's the little squirt that gives up easy so mel needs to backtrack
and that'll be your mel
can i retrain with added dataset?
only with a new session
but same epochs? or i have to start again
because it's a whole different dataset
ok - let me reframe the earlier question because I was trying to get a feel for how much labor is involved because I'd honestly be willing to pay somebody to do the tensorboard work and do a couple of fusions for me.
Ayo? @crimson wraith level 4 !!! 
start again
rip
it is not hard, you're just impatient
it takes a while to learn something
but yeah do a
Indeed.
Hi, I'm experiencing a lot of lag when synthesizing the voice... it's practically unreadable, I have a xeon2670v3 and an Rx5500XT
Fascinating. I didn't even know that was a thing. I'll check it out - also checking out learning tensorboard.
-colab
- Applio, by IA Hispano Google Colab
- RVC Disconnected, by Kit Lemonfoot Google Colab
- RVC Mainline, by Hina Google Colab
- AICoverGen-WebUI, modded by Hina Google Colab
- AICoverGen-NoWebUI [English], by Ardha, fixed by Eddy, Hina and Gdr Google Colab
- AICoverGen-NoWebUI [Spanish], credits to Eddy, Hina and Gdr for translating and fixing Google Colab
- Ilaria RVC, by thestingerx Google Colab
- EasyGUI, by rejects Google Colab
- UVR5 NO UI for Google Colab, by Eddy Google Colab
- 🆕 UVR5 UI, by Eddy Google Colab
-kaggle
- 🆕 UVR5 NO UI by Eddy
- How to use RVC Mainline Kaggle by Cauthess
- 🍏 Applio RVC by IA Hispano
- ✨ RVC Mainline by Hina
hey
please help me, how to run applio via router IP?
because applio doesn't working with Wi-Fi on mobile device
mods you are offline??
retrain from the beginning (preprocess)
vpn or try on pc/laptop
I use local version
how is ethernet or router wifi?
router wifi, local link works, but blocking on a wi-fi IP
then VPN as I said @tacit badger
I can change IP address in Applio or not?
what's the point? if it's 192.168.* it's local network address so it shouldnt get blocked by the ISP
(not mentioning possibly bricked firewall/network configurations)
- Applio, by IA Hispano Google Colab
- RVC Disconnected, by Kit Lemonfoot Google Colab
- RVC Mainline, by Hina Google Colab
- AICoverGen-WebUI, modded by Hina Google Colab
- AICoverGen-NoWebUI [English], by Ardha, fixed by Eddy, Hina and Gdr Google Colab
- AICoverGen-NoWebUI [Spanish], credits to Eddy, Hina and Gdr for translating and fixing Google Colab
- Ilaria RVC, by thestingerx Google Colab
- EasyGUI, by rejects Google Colab
- UVR5 NO UI for Google Colab, by Eddy Google Colab
- 🆕 UVR5 UI, by Eddy Google Colab
hey it's giving me an error and idk what i did wrong
halo i need help rvc app suddenly sounds like a broken megaphone when changing its voice
please specify the error and what you are using
what error and what are you using
I tried downloading the index file for a model I saved in w-okada, but it ended up being a 0-byte file. Now, I can't delete it. I've fixed all the other issues I had, but this one is still bugging me. do you have a fix for it?
why does rvc wont detect my gpu
my rvc trained 2 and a half epochs and has now just stopped
whitch one should i use? D_2333333.pth , G-2333333.pth or lobothexile.pth? D is 827.4MB, G is 417.4MB and lobothexile is 52.7MB
I use RVCDisconnected for training
the full pth file (not g/d)
the g and d file is for training more
:p
oh got it thanks!
yo i need help
uhhh, i dont know much about wokada, its better u ask in #🔍│help-w-okada or #1192011222023950368
are you talking about realtime rvc for calls (wokada) ? if so the best channel to ask help in would be #1192011222023950368 or #🔍│help-w-okada
pls can anyone can train voice model
hello i want to make an ai that plays war thunder tank at a decent level how can i achive that
I already wrote but no one helps
you can learn how to do it in https://docs.aihub.wtf ( u can do it locally if u got a good pc, else in cloud)
Or, do a free / paid request in #1159289738314919936
Or, commission a model master #1191429836321849435
Last update: Mar 10, 2024
Sorry but i can't help you with that, hopefully an helper comes soon to help you
-colab
- Applio, by IA Hispano Google Colab
- RVC Disconnected, by Kit Lemonfoot Google Colab
- RVC Mainline, by Hina Google Colab
- AICoverGen-WebUI, modded by Hina Google Colab
- AICoverGen-NoWebUI [English], by Ardha, fixed by Eddy, Hina and Gdr Google Colab
- AICoverGen-NoWebUI [Spanish], credits to Eddy, Hina and Gdr for translating and fixing Google Colab
- Ilaria RVC, by thestingerx Google Colab
- EasyGUI, by rejects Google Colab
- UVR5 NO UI for Google Colab, by Eddy Google Colab
- 🆕 UVR5 UI, by Eddy Google Colab
Hi, is there amy way to make a multispeaker RVC model with a database looking like this?
Model
└───D
├───.wav
└───C
├───.wav
└───B
├───.wav
└───A
├───.wav
Well you can either just dump all the audio files in the same folder and train like that (actually works, but I would suggest, in the case of 3 voices, two should have the same length of dataset and the one you want most dominant should have like 2-3 minutes more)
Or you train separate models, and then merge them with percentages you can set
Ah, I see, It wouldn't work out for me ig
no
TypeError: cannot unpack non-iterable ValueError object
is there any other way i can try model blending
@proven hill does the ilaria colab support model blending
ilaria mainline does
oh this colab then?
https://colab.research.google.com/drive/1mHKTGH5e3SAyDSBss1KtiYRbDdQzwSMs
Ayo? @brittle wing level 11 !!! 
yes
perfect thank you!
no problem
guys my voice changer has some robot sound
Ayo? @tulip creek level 1 !!! 
what i need to do
-colab
- Applio, by IA Hispano Google Colab
- RVC Disconnected, by Kit Lemonfoot Google Colab
- RVC Mainline, by Hina Google Colab
- AICoverGen-WebUI, modded by Hina Google Colab
- AICoverGen-NoWebUI [English], by Ardha, fixed by Eddy, Hina and Gdr Google Colab
- AICoverGen-NoWebUI [Spanish], credits to Eddy, Hina and Gdr for translating and fixing Google Colab
- Ilaria RVC, by thestingerx Google Colab
- EasyGUI, by rejects Google Colab
- UVR5 NO UI for Google Colab, by Eddy Google Colab
- 🆕 UVR5 UI, by Eddy Google Colab
send the screenshot of okada
wait i have voice changer demo
should work as intended
i personally prefer it and its more supported
lol
lower the chunk to 192
also use sup2 instead of echo
and set the index to 0
somethings its wonky
you should read this guide
EOL - No further Updates
Github - Blanc-dot
Discord User ID - 824922747423031359
Despite being end of life, most if not all information has not really changed, so should be very accurate until actual new stuff comes out.
Other Links
Antasma's Local Error Fixes
Antasma's Colab guide
Sushi's useful...
i know how to fix that
ho
open your task manager
go to details
find this process
set the priority to high
and also set the affinity to only one of these processors
that's what fixed it for me.
thx
you will have to redo it after every pc restart
there is a way to change it permanently
REALLY WORKS
you have to do it with process lasso https://bitsum.com/
Guys does anyone know what character has the voice to do commentary in yt vids
bro someone keeps sending fricking 18+ images and stuff and deleting them like 0.2 seconds later
in this server
who is it
who?
hu
can anyone ban the user?
@red kayak
is that him
no
oh
i think they stopped now
but i swear to god there were flashes of an invite to a server with like a peach emoji and the 18+ sign on it
ok but make sure this does not become a threat......
yeah
wait, wdym threat
i cant do anyhting since im not a moderator, only a member
-rvc
- How to use RVC Mainline Colab by Cauthess
- Full AI Voice Model Training Guide (local) by Christopher Villanueva
- AICoverGen Colab Guide by Eddy (Spanish Helper)
- Create a model with RVC disconnected (colab) by Angetyde
Already banned
Hi mates! For you, what's the best rvc version for training with apple silicon
no one
Don't think anyone's pulled that off tbf
But it might be a thing. Maybe just needs to be tweaked
Fck really ?
you have to train it using colab
Same if on Windows, should be possible on Linux
Been a while since I haven't trained nothing
Wait really ?
So like in one year no one pulled out a version for apple silicon or for amd gpus on windows ??
It can be possible once Zluda support is introduced (and if your GPU can work with ROCm without additional manipulations). There's a PR for that in Applio
https://github.com/IAHispano/Applio/pull/504
But until then, it's not possible. No one wants to mess with training on DirectML and there's a handful of people who even use RVC on mac
Idk what ROCm is but it sounds cool
A backend for AMD. And it may sound cool, until you figure out your GPU is not supported out of the box
Should be working
https://rocm.docs.amd.com/en/docs-5.5.1/release/windows_support.html
Wdym by "tweaked"
Yeah ok sure but what tweak should I do to get it working ?
why is the voice changer so delayed
What’s your GPU?
Ok so I just need to wait the merge of the pull request ?
Ayo? @fallow kindle level 1 !!! 
stfu
https://github.com/deiteris/voice-changer since you have a AMD GPU I would recommend using the optimized fork
Which should have less delay
thanks
which one do i download
dml
Hi ! i have 2 separate disks C (4gb space) and E ( 100gb space) so when i try to train model it takes from the C disk space is there any way to make it take from the E disk space?
On which disk is RVC installed?
C or E?
E
everything is on the E disk i don't know why it takes from the C disk
That's pretty weird.
Have you set "Save only the latest ckpt" to no?
yes, should i put it to yes ?
-applio
Ayo? @wheat hedge level 1 !!! 
- 🍏 Applio, by IA Hispano GitHub
- Mangio-RVC-Fork, by Mangio621 Huggingface
- RVC Studio, by SayanoAI Huggingface
- AICoverGen, by SociallyIneptWeeb GitHub
- Replay, by Replay Team Website
- Original RVC, by the RVC-Project team GitHub
- GPT-SoVITS, by RVC-Boss GitHub
Credits to Faze Masta and Antasma for compiling these links.
@odd shale i will try to delete all the rvc files and download it again maybe it will solve the problem
Try doing so then.
Does AICoverGen-WebUI still work?
Ayo? @brittle wing level 4 !!! 
Dont see the Link I usually click to open the RVC
How would I go about continuing training of a model in Applio after it has stopped?
Are you trying to train or make covers
guys i need help the voice is not working
hi! i just tried installing the voice model and
it's just not working at all 😭
it loads but then does nothing
why does it say 0 0 0
-colab
- Applio, by IA Hispano Google Colab
- RVC Disconnected, by Kit Lemonfoot Google Colab
- RVC Mainline, by Hina Google Colab
- AICoverGen-WebUI, modded by Hina Google Colab
- AICoverGen-NoWebUI [English], by Ardha, fixed by Eddy, Hina and Gdr Google Colab
- AICoverGen-NoWebUI [Spanish], credits to Eddy, Hina and Gdr for translating and fixing Google Colab
- Ilaria RVC, by thestingerx Google Colab
- EasyGUI, by rejects Google Colab
- UVR5 NO UI for Google Colab, by Eddy Google Colab
- 🆕 UVR5 UI, by Eddy Google Colab
!howtoask
How To Troubleshoot 
- Don't simply mention your issue, like "
my rvc is not working". - Describe the step you are on, what you're trying to do, the RVC you're using, a screenshot, etc.
- The more context, the better.
- Don't be desperate. You can ping a Helper, but if they ignore, they aren't available/don't know the answer.
- It's okay if you're frustrated, but don't take it into this server.
- Don't DM without prior consent.
- Don't ask for every little instruction. Put your own effort & test things by yourself.
- Don't ask to ask.
- Check if your answer is a Google search away/on our guides website.
Should I keep training or is it over trained already?
can someone help me download rvc n how to set it up
i keep freeezing using wasapi, i have same HZ range set, anyone else experience this?
-colab
- Applio, by IA Hispano Google Colab
- RVC Disconnected, by Kit Lemonfoot Google Colab
- RVC Mainline, by Hina Google Colab
- AICoverGen-WebUI, modded by Hina Google Colab
- AICoverGen-NoWebUI [English], by Ardha, fixed by Eddy, Hina and Gdr Google Colab
- AICoverGen-NoWebUI [Spanish], credits to Eddy, Hina and Gdr for translating and fixing Google Colab
- Ilaria RVC, by thestingerx Google Colab
- EasyGUI, by rejects Google Colab
- UVR5 NO UI for Google Colab, by Eddy Google Colab
- 🆕 UVR5 UI, by Eddy Google Colab
Is AMD GPU’s like RX 7900 XTX available for Ai Training Locally in Windows?
i have a rtx 4080 which rvc should i use
no
for the voice changer?
logs
added index
Thank you
the added.index and trained.index file have any different ? :'∆
use the added index
for both voice changing and training
Ayo? @twin pagoda level 1 !!! 
for the voice changer follow this guide: https://rentry.co/VoiceChangerGuide
EOL - No further Updates
Github - Blanc-dot
Discord User ID - 824922747423031359
Despite being end of life, most if not all information has not really changed, so should be very accurate until actual new stuff comes out.
Other Links
Antasma's Local Error Fixes
Antasma's Colab guide
Sushi's useful...
hello i have problem with RVC I can't make it work with discord, can't hear me, plase help me;c
https://cdn.discordapp.com/attachments/1264933226082730055/1264933226841899060/21312321Untitled.png?ex=669fac5e&is=669e5ade&hm=dcf8ca0bfd53c3548656002f26df8a8a1456316b10dfea0657c8c9db458e7bb2&
https://cdn.discordapp.com/attachments/1264933226082730055/1264933227127378053/rvcrvc.png?ex=669fac5e&is=669e5ade&hm=ad22b2f29e862f61886def4c136a47d444e1fe7c0e1bfd87c3eada9956163f77&
https://cdn.discordapp.com/attachments/1264933226082730055/1264933226623930471/21312321Untitled.png?ex=669fac5e&is=669e5ade&hm=89588f8c1b268fddf79ef480279ff64159e0bfbab691186eb29627e20f39a402&
https://cdn.discordapp.com/attachments/1264933226082730055/1264933226263347250/123213213.png?ex=669fac5e&is=669e5ade&hm=5d475556cf2426cc05707abc9ddb6c4f1e39b1d6a4003085bc8361f46366a81a&
Do you have your mic set on discord to the audio cable?
-colab
- Applio, by IA Hispano Google Colab
- RVC Disconnected, by Kit Lemonfoot Google Colab
- RVC Mainline, by Hina Google Colab
- AICoverGen-WebUI, modded by Hina Google Colab
- AICoverGen-NoWebUI [English], by Ardha, fixed by Eddy, Hina and Gdr Google Colab
- AICoverGen-NoWebUI [Spanish], credits to Eddy, Hina and Gdr for translating and fixing Google Colab
- Ilaria RVC, by thestingerx Google Colab
- EasyGUI, by rejects Google Colab
- UVR5 NO UI for Google Colab, by Eddy Google Colab
- 🆕 UVR5 UI, by Eddy Google Colab
is there a reccomended website to download music?
EOL - No further Updates
Github - Blanc-dot
Discord User ID - 824922747423031359
Despite being end of life, most if not all information has not really changed, so should be very accurate until actual new stuff comes out.
Other Links
Antasma's Local Error Fixes
Antasma's Colab guide
Sushi's useful...
thank you
what is the command to see what is the good settings for amd and nvidia gpus?
Ayo? @brittle wing level 3 !!! 
⠀
Settings for Nvidia GPUs 
F0 Det.: rmvpe (suggested for all series)
RTX 40-series: 80-96 chunk | +16384 extra
RTX 30-series: 96-112 chunk | +16384 extra
RTX 20-series: 112-128 chunk | +16384 extra
GTX 16-series: 128-192 chunk | +8192 extra
GTX 10-series: 128-192 chunk | +8192 extra
Advanced Settings
Protocol : Sio or Rest
Crossfade: 4096 start 0.2 end 0.8
Trancate: 300
Silencefront: Off
Protect: 0.5
RVC Quality: Low
⠀
Ayo? @scarlet folio level 1 !!! 
why is there a guide for okada when this is rvc
is this okada?
i meant this?
Ayo? @brittle wing level 1 !!! 

AI HUB Docs