#✨│ai-help
1 messages · Page 167 of 1
both laptop and pc
3080 laptop and 4070 pc
sure ask alot when you could just tell me wheres the link
Reviving in the future, will change install instructions to be "manual" build (for nvidia at least as its infinitely better performance)
Github - Blanc-dot
Discord User ID - https://discord.com/users/824922747423031359
Despite being end of life, most if not all information has not reall...
This should have the download
And settings, setup, etc
Or if you want better performance you can use the optimized fork but you won’t see much difference since your GPU is already pretty good https://rentry.co/bettervoicechangerguide
Every step in here is explained on the GitHub aswell which you can find here. It might be more convenient for some people, but others prefer to not read much, which this guide is for.
Guide style is in the same as Blanc-dot's guide for simplicity and familiarity sake
Why should you use this versi...
yes this one is the one i like more
Ayo? @junior canopy level 1 !!! 
UVR BVEでもそれができる
For some reason it works fine but my friend says he could still hear me and the ai voice, is there any fix to that
hearing both together? if youre using a kind of voicemeeter, check the audio routing configurations
Ty
Ayo? @royal pike level 1 !!! 
i got a 15 min dataset what batch size should i use
Ayo? @paper dock level 12 !!! 
What model of Riala RVC vocal remover should I choose? Also I would like to know the functions of the 7 ones
If you want to be fast then 10-12
okay, are you talking about ilaria RVC or still the UVR? and which 7 functions do you mean?
イラリア RVC について話しているのですか、それともまだ UVR ですか? また、どの 7 つの機能について言っているのですか?
イラリアrvcのUVR機能のことです
えっと、UVRアプリは次のようになります https://huggingface.co/spaces/TheStinger/UVR5_UI
ごめんなさいイラリアのvocal remover 機能のことです
イラリア内のvocal removerは良くないので、このUVR UIは新しくて優れています
https://huggingface.co/spaces/TheStinger/UVR5_UI
Hugging ai なんか連続で使ったらgpuが何とか出てきたんですけどどうすればいいですか?
GPUの使用制限に達したと表示されましたか?
多分そうです
Ayo? @final sentinel level 4 !!! 
赤文字で
I'll just leave it here lol
#🔍│help-w-okada message
im using rvc mainline and i noticed it doesn't have 32k option
which alterantive should i use
for the option lol
そうだね…
2つのオプションがあります:
- UVR Colabノートブックでは、毎日最大4時間の無料GPUセッションが提供されますが、最初に必要なパッケージをインストールするのに数分かかる場合があります。
https://colab.research.google.com/drive/16Q44VBJiIrXOgTINztVDVeb0XKhLKHwl - PC/ラップトップを必要とするローカル UVR インストール。CPU モードで実行できますが、Nvidia GTX/RTX 10 シリーズ以降を使用すると、より高速に実行できます。
https://github.com/Anjok07/ultimatevocalremovergui
BS Roformerモデル用のRoformerパッチをインストールできます
https://github.com/TRvlvr/model_repo/releases/download/uvr_update_patches/UVR_Patch_4_14_24_18_7_BETA_full_Roformer.exe
I'd like to do the Uvr collaboration notebook, what should I do first?
次の画像に示すように、必要なセルを順番にクリックします。何かプロンプトが表示されたら、確認します。
will this work on CPU Basic? The ZeroGPU requires a Huggingface subscription
ごめんなさい、このColabの方が新しくて良いかもしれません
https://github.com/jarredou/MVSEP-MDX23-Colab_v2
-hf
- 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
ただし、エラーが発生した場合、私は試したことがないのでわかりません。また、私が使用した他の Colab でもパッケージ競合の問題がいくつか発生し、最近修正されました。ただし、MDX23 Colab についてはよくわかりません。
I installed it, what do I do?
You're supposed to toggle it back and forth to v1 and v2. It is not v2 if you don't see 32k
hmm?
that's how it is on Applio too
ig, at least thats how I made sense of it. It threw me a tensorboard error if I just train as is because the pretrains were 32k
so the target sample rate needed to be 32k too
that MDX23 Colab may seem complicated, but by the way there is another Colab you can try on
https://colab.research.google.com/github/Eddycrack864/UVR5-UI/blob/main/UVR_UI.ipynb#scrollTo=oXYEyNjdWaMm
you can check more guide here
https://docs.google.com/document/d/17fjNvJzj8ZGSer7c7OFe_CNfUKbAxEh_OBv94ZdRG5c/edit#heading=h.k3vca4e9ena8
MDX23 Colabは複雑に思えるかもしれませんが、別のColabを試すこともできます
https://colab.research.google.com/github/Eddycrack864/UVR5-UI/blob/main/UVR_UI.ipynb#scrollTo=oXYEyNjdWaMm
詳しいガイドはこちらをご覧ください
https://docs.google.com/document/d/17fjNvJzj8ZGSer7c7OFe_CNfUKbAxEh_OBv94ZdRG5c/edit#heading=h.k3vca4e9ena8
edit 19.07.24 deton24’s Instrumental and vocal & stems separation & mastering guide (UVR 5 GUI - VR/MDX-Net/MDX23C/Demucs 1-4, BS/Mel-Roformer MVSEP-MDX23-Colab/KaraFan/drumsep/LarsNet/x-minus.pro/Ripple/GSEP/Dango.ai/Audioshake/Music.ai) General reading advice | Discord | Table of content (or ...
not to be paranoid about it but I have to assume that is V1 by default
we wouldn't have two folders otherwise
anyways just toggle it
I'm trying to convert a 30 minute audio clip and both Applio and RVC v2 on Huggingface keep throwing errors. Can anyone help or tell me why I'm getting errors?
Ayo? @shut fog level 1 !!! 
it would eat up huge memory, so you should split it and convert each separately, or you can try enable "Split Audio" in Applio
Ayo? @knotty moth level 22 !!! 
Alright I was already using the split option so I guess I'll split it manually thanks
I installed it, what do I do?
umm, if you need more guidance, can I add you please? so you can DM me with screenshot of some steps you are stuck on.
えーと、さらに詳しい説明が必要な場合は、追加していただけますか?行き詰まっている手順のスクリーンショットを添えて、DM で送ってください。

although your 4070 is more than enough with correct settings, you can try this https://github.com/deiteris/voice-changer
and follow the guide as well https://rentry.co/bettervoicechangerguide
リアルタイムボイスチェンジャー Realtime Voice Changer. Contribute to deiteris/voice-changer development by creating an account on GitHub.
Every step in here is explained on the GitHub aswell which you can find here. It might be more convenient for some people, but others prefer to not read much, which this guide is for.
Guide style is in the same as Blanc-dot's guide for simplicity and familiarity sake
Why should you use this versi...
can i get help in spanish?
don't ask to ask, but if you don't mind my borked translations, just ask it 
umm guys ?
use gemini /hj
- 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
- UVR5 NO UI for Google Colab, by Eddy Google Colab
- UVR5 UI, by Eddy Google Colab
#📰│dev-updates message read dev updates please ! ! !
oh
ye rip easygui
-local
- 🍏 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.
How do you continue training models with rvc disconnected?
Ayo? @balmy reef level 2 !!! 
Let's say a model is on 600 epochs and one wants 900. How can one do it?
need the G and D paths which are in logs to continue training
赤文字のものは何を言っているのでしょうか?それとその赤文字のものをできる限り出なくする方法はなんでしょうか?
What do the red-letter things say? And what is the way to make those red-lettered things disappear as much as possible?
あそこで言ったような感じです #✨│ai-help message
代わりにこれをお勧めします https://mvsep.com/
Does a resume session on training usually take minutes rather than starting from scratch and it takes a bit?
In other words, rather than 1-2 hours, it takes roughly 30 min to finish resumed rvc training.
log in to the site to have more data
you can resume training if you have kept/saved the preprocessed logs folder and G & D files, etc within it and load them all later, so wdym
Yes, I had those files but the resume session took minutes rather than hours.
then why not 
That's pretty cool. I was expecting a resume session to take the same amount of time.

There seems to be a lot of different RVCs to use for training. What do you guys think sound the best for local training? I searched for people reviewing the different RVCs and I couldn't find anyone talking about their differences
There isn't any difference in quality tbh. It's all about stability, ease of use and extra features you get on each RVC fork (fork is what we call the different RVCs)
I'd recommend using Mainline tbh
Last update: Mar 8, 2024
Thank you very very much for answering me and for the recommendation ❤️
Ayo? @grave frigate level 1 !!! 
You're welcome :)
does any one know how i use my graphics card and not the cpu on the new update
What’s your GPU
How do you use an updated trained rvc?
I keep loading the drive link and it says it's successfully uploaded but when I go to load the model I don't see it.
Do I need to recreate another zip file?
i got a rtx2060 but its fine i have had this sofware before but they made some bad updates and i dont what to do now sadly
Maybe drive links r screwed up... I'd suggest using Huggingface for model uploads instead
Does it automatically update the original zip file? I am using Huggingface to load the models but had an easier time generating a link with Google Drive. 😅
When you first train the model there's always the index file. I resumed training for 2 rvcs earlier but I didn't see one, I assumed it updated the original zipped file with the index and pth.
if you use the Huggingface export model step yes
if you use rvc Disconnected that is
I did use rvc Disconnected! However, I went to the original zip file on drive and it says last updated was a few days ago, whoops lol
Unless rvc Disconnected didn't show any update and it already did?
I can't download Gosegu rvc, what is it?
gosegu rvc ?
When I click download, it says, "Invalid username or password."
what are you downloading? dont download things off youtube tutorials
I'd like to download gosegu rvc
Ayo? @restive ocean level 1 !!! 
send where u got that download link
is it a voice model that you are talking about?
it shouldnt ask you for a password
send exactly where u found whatever u trying to download
yes
send the voice model download link
Thank you
huh?
i asked you to send the download link of the voice model to check it
the model repository has been deleted or privated
thats why
you need to try another model
is there somewhere u found that link?
no
alright, just try another model as u cant get that one anymore
ok, thank you
yw
I ran every step in the document except Index Training
I resumed training on an already trained RVC
I'm gonna try again. Hang on....
Ayo? @balmy reef level 3 !!! 
hello can anyone help me. idk how to set up "crepe" in the f0 det. settings. it says "N/A" what do i have to do to enable it
I assume you use wokada
whats your gpu
crepe is not the go to f0 det these days, its fine
i have an rx 6950 xt
i recommend you use the modified wokada, its better for amd gpus
https://rentry.co/BetterVoiceChangerGuide
there you use rmvpe_onnx
Every step in here is explained on the GitHub aswell which you can find here. It might be more convenient for some people, but others prefer to not read much, which this guide is for.
Guide style is in the same as Blanc-dot's guide for simplicity and familiarity sake
Why should you use this versi...
the only two options i have are "crepe_tiny" and "rmvep_onxx"
i'm using the one you're mentioning. Still it sounds reaaaaally laggy
I dont think youre using it since u only mentioned 2 options while it has more
laggy is because you probably didnt export to onnx which is something you have to always do on amd gpus
on the one i linked you dont have to do that
if you did then could be chunk and extra setting
can anyone help me? im searching for index file in voice models but every have only one file which is pth file
its in logs folder
logs folder?
where is it?
Ayo? @brittle wing level 1 !!! 
who can i contact for this https://huggingface.co/spaces/r3gm/Advanced-RVC-Inference as the downloader is not working i keep getting error
just use ilaria rvc
where and how to find logs folder?
i pressed train model so do i just wait cuz there is not stop trianing button
Ayo? @true hatch level 2 !!! 
im sure i did everything right
-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
- UVR5 NO UI for Google Colab, by Eddy Google Colab
- UVR5 UI, by Eddy Google Colab
but its not what im looking for like that ui is
does voice changer need index file to work on discord?
- 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
- UVR5 NO UI for Google Colab, by Eddy Google Colab
- UVR5 UI, by Eddy Google Colab
theres a folder called logs in the main directory
in the files of voice changer?
then why it wont work?
what no why would it be in voice changer , i thought you trained a voice model and asked where it is. theres no index file there, if you download a model it can come with index but sometimes doesnt
explain what doesnt work
its not because of index 100%
im not training anything
im using this
whats your GPU name?
Radeon RX 570 Series
with that gpu i recommend you download the modified voice changer, it runs a lot better on radeon amd gpus
https://rentry.co/BetterVoiceChangerGuide
follow amd steps and virtual cable if u dont have it yet
Every step in here is explained on the GitHub aswell which you can find here. It might be more convenient for some people, but others prefer to not read much, which this guide is for.
Guide style is in the same as Blanc-dot's guide for simplicity and familiarity sake
Why should you use this versi...
the normal wokada that you have for rx 570 i dont recommemd
alr, but how i should use it?
is it on yt?
no, the guide i linked is rlly straight forward non complicated youll be fine
it should be showing as website?
Ayo? @brittle wing level 2 !!! 
yes
it sounds normally, but discord dont catch that voice
can you screenshot the voice changer
and what you put as input output on discord
should i change output to the same as voice changer?
on discord INPUT should be virtual cable
also i recommend rmvpe_onnx instead of crepe
also chunk im sure you should be able to run a lower number, idk which but maybe 300, maybe lower
lower means less day on that btw
less delay*
and for voice changer?
the normal one?
i mean
default
voice changer is correct already
there its on output
it still dont work
im pressing start and still nothing
what kind of virtual cable did you install?
did you download VAC lite from the guide?
no, im gonna install it
i dont recognize the name from your screenshots so could be that its not working if its different
alright
what is that version you are using? my UI looks different
the one that he sended me
Probably the deiteris fork
the modified version ye
https://rentry.co/bettervoicechangerguide
Every step in here is explained on the GitHub aswell which you can find here. It might be more convenient for some people, but others prefer to not read much, which this guide is for.
Guide style is in the same as Blanc-dot's guide for simplicity and familiarity sake
Why should you use this versi...
its running as admin by itself
still nothing
alr
i fixed it
i set up everything correctly
still nothing
btw did you ask for raven approval to copy paste some stuff?
removed the copy paste and blanc dot is mentioned on the top, why
wouldve been cool to message her
Ayo? @brittle wing level 2 !!! 
What batch size would you guys recommend for a 23 minute long dataset
Ok? can do ig can u send discord tag instead of id
baka.aqua
Does the model name have any spaces
same download link as yesterday?
Yeah
i just re used the same download and downloaded it as 'REAL-frankocean' and it works fine, are you sure you dont miss a step or something when u download a model?
Nope…
I’m trying to convert a 50 sec recording
Is that why it’s causing a problem?
Nope, i think its someway u download the voice models but i dont really understand as everytime i download it to test they work fine
you follow the guide right?
whats happening
idk but everytime the guy downloads the model, it always errors out, but when i download with the same link, it works
theres clearly some mistake being made
RVC = Retrieval-based-Voice-Conversion
maybe u could try to ask him what he does step by step cus idk its the 2nd time
thats the error i get 😭
Ayo? @nimble tangle level 2 !!! 
Is there a guide to the EasyGUI because I honestly just don't know what I'm doing. I keep getting an error that says there is no "faiss" thing but I don't know what that is
If you are talking about the easygui google colab, its broken #📰│dev-updates message
OH SORRY That must be it. Thank you!!
Dw, i'd recommend you to use the alternatives as the easygui google colab is broken since days
Okay! Is the an alternative you recommend for beginners?
Ayo? @brittle wing level 4 !!! 
Well, there's guides for them, do you need inference (use models) or train (make models) ?
Train :D
use rvc disconnected hun
what
lmao, i hang out here sometimes
HAHA I had some issues with the Caseoh model and you helped me out via DMs
I just remembered it
You can use:
- RVCDISCONNECTED (has no ui)
- Applio Colab (has ui)
Its personal choice, i personally used rvcdisconnected (ui means interface btw)
Last update: Mar 8, 2024
Last update: June 15, 2024
OH GOD
I REMEMBER
I'm going to use RVC Disconnected, thank you so much!!
OMG YOU REMEMEBER HAHA I totally gave up on that
your welcome
things got easier with ilaria rvc zero
Oh really? That's great!
Well I'll definitely use it then! I am on a completely different mission now, so hopefully it turns out better. I never have luck with these things
let me know if i can help
Thank you I will!!
Hey how would one work the HuggingFace export model cell on rvc2 Disconnected? Before that I noticed an issue with the "Train Index" cell and how the output is "no feature todo" when it should say "all feature done". I am resuming training on a model everything and works okay up until Huggingface export. Help?
(please ping me, thank you 🙂 )
How do i make a ai model on ios easy?
What is the best setting for Haggingface's UVR to get the cleanest vocals?
hey i have a problem with the cable input when i try to use it on dc it sounds so robotic it makes like weird sounds but when i use it normally with my normal headphones it all sounds clear how can i fix it?
Use VAC LITE instead of VB audio
-rt
This interaction has expired, use the command
/guides realtimeif you wish to see it again.
-realtime
This interaction has expired, use the command
/guides realtimeif you wish to see it again.
Reviving in the future, will change install instructions to be "manual" build (for nvidia at least as its infinitely better performance)
Github - Blanc-dot
Discord User ID - https://discord.com/users/824922747423031359
Despite being end of life, most if not all information has not reall...
AEUGHHHH
no image perms
help
i need
trying to bopot up the thing with ngrok and it just says this'
/content/voice-changer/server
ModuleNotFoundError Traceback (most recent call last)
<ipython-input-2-ca461f55b6bc> in <cell line: 24>()
22 get_ipython().run_line_magic('cd', '/content/voice-changer/server')
23
---> 24 from pyngrok import conf, ngrok
25 MyConfig = conf.PyngrokConfig()
26 MyConfig.auth_token = Token
ModuleNotFoundError: No module named 'pyngrok'
NOTE: If your import is failing due to a missing package, you can
manually install dependencies using either !pip or !apt.
To view examples of installing some common dependencies, click the
"Open Examples" button below.
---------------------------------------------------------------------------\
ixed it
hey
i think i downloaded the wrong version
can someone tell me what is the voice changer version i have to download for this
what you can do is cope
integrated graphic card wont take you anywhere
lol u are not helping by the way
im not an helper
what is that
the thing you wanted
Every step in here is explained on the GitHub aswell which you can find here. It might be more convenient for some people, but others prefer to not read much, which this guide is for.
Guide style is in the same as Blanc-dot's guide for simplicity and familiarity sake.
Why should you use this vers...
follow this guide
i mean the whole program
thats the whole program
well imma try
Ayo? @nimble tangle level 3 !!! 
thanks tho
np
yo
im generating a voice model but how do i stop it?
the guide says to press the stop training button where the train model button is but i don't see a stop training button
close out rvc and make sure these are enabled
Ayo? @thin stump level 34 !!! 
next time
k
how do i use my trained model in w-okada
like how can i extract my trained model to a .pth file
to use in w-okada
What is a good epoch level because like my runtime disconnects at Epoch 150 every time
Like it can only reach 150 before like going "Okay you're done you have to reconnect and restart"
psss Have you heard of kaggle?
-kaggle
- UVR5 NO UI by Eddy
- How to use RVC Mainline Kaggle by Cauthess
- 🍏 Applio RVC by IA Hispano
- ✨ RVC Mainline by Hina
30 hours of GPU weekly. Here's 8 hours of me sleeping lmao
WOAH :O
but you have to manage the drive once in a while to recycle epochs that you don't want. It's all in the guide so dw
I'm so bad at this I am looking at that and I'm very overwhelmed and clueless 
HAHAHA
I was thinking of making a speedrun video once I get a dataset ready
. Muscle memory goes brr
If you want I would love that
Hey how would one work the HuggingFace export model cell on rvc2 Disconnected? Before that I noticed an issue with the "Train Index" cell and how the output is "no feature todo" when it should say "all feature done". I am resuming training on a model everything and works okay up until Huggingface export. Help? (please ping me, thank you 🙂 )
Hi,
I had the model maker role and made the "Thomas the Tank Engine (Building the New Line)" voice model in the old Ai Hub discord server. I am wondering if it's possible to get my perms back. Thanks
Hi! Um, is it normal for it to still be running but say Draft? Should I start over?
no its ok, theres so much gpu so you cant mess up
like even if you idle on the page
Ohh okay I was just curious. It was taking a while so I thought maybe I had messed up somehow LMAO The guide said the second one would take a while but not the fourth one so I got worried
Thank you!!
-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
-hf
- 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
jk jk, but check your audio files
to see if its in wav format
Google colab always says to me failed
How long does the fourth cell take? It's been 56 minutes 😕
what fourth cell? Because I only use !wget since its more convenient for me
to download the pretrains
It said execute all the cells manually so I was doing that and I've counted 5. My bad I'm on the fifth cell not the fourth
oh, have you authenticated your ngrok already?
I think so
you should double check since it'll give you the three links afterwards https://dashboard.ngrok.com/get-started/your-authtoken
ngrok is the fastest way to put anything on the internet with a single command.
Oh I might not have then hold on
I copied the authentication code but where do I put it?
the fifth or final cell
i call it the final cell in the guide
No I was wanting to authenticate my Authtoken and I couldn't figure out how
There wasn't a run cell option or anything so I assumed you put it somewhere
oh i see, idk how I did it so ig you do have to use the command line
Yeah but where do you use it? Do you make a new code on the Kaggle page or is there somewhere else to copy paste to use it?
did you open cmd on your taskbar and enter this in
with your auth token
No let me try that
then after that has been authenticated
oh ;p try this too
get it from here
https://dashboard.ngrok.com/get-started/your-authtoken
ngrok is the fastest way to put anything on the internet with a single command.
I have the token I'm just trying to authenticate it
Ayo? @brittle wing level 5 !!! 
And I have no idea how
I'll update the guide if you do figure it out 🥹 . I don't have another account to try it out
i gtg
Alright have fun!
well, the point is just to copy the authtoken, paste to the cell and run it
or did you have sort of "authentication" error when trying to do that?
bruh
at least it should immediately show the three links (rvc, tensorboard, imjoy file manager)
It is
then open all of those
can someone help me, i selected the audio model and audio but it says "please select a model input and input audio file" when i try running it
Jk jk? Did you message me twice lol? Anyway, yeah they were originally in .wav format I'm resuming training on a zip file with a pth and index file inside of it because they're already trained on I think 500e
I was just saying use kaggle too because I don't train on rvc disconnected anymore. You can't retrain a pth file and index if they're already created by the way. You can resume training if its your model though
Last update: Mar 8, 2024
-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
I'm trying the RVC Mainline locally but the tutorial never says where I can open the TensorBoard and I can't find it in the files. Can someone help me?
extra step, ik so annoying https://docs.aihub.wtf/rvc/resources/epochs--tensorboard/#installing-opening
Last update: Feb 10, 2024
thank you a lot! ❤️
Hmmm alright....well I do have a model and training is successful but idk how to re-use it in a hugging face repo lol
What model of haggingface uvr can separate vocal and inst most beautifully?
i never used that cell before actually. If you just want do inferencing then https://docs.google.com/document/d/1YbXcLFPaGjhOdG5NFkK3QrucCEpHZBwFUxkeMO8aB18/edit
Table Of Contents Introduction (with link) Model Loader (Download & Upload) Inference (use RVC AI Voice Models) Ilaria TTS Settings (Inference) Introduction (with link) Ilaria RVC Zero, is an RVC (Retrieval-based Voice Conversion) Fork made by Ilaria & mikus, running only on Hugging Face Spac...
How do I fix when my friend talk it also record it
BS Roformer 1296 or 1297
there's no huggingspace where we can do training if thats what you're asking. I'm slightly confused on what you're trying to do
so @balmy reef
On rvcDisconnected there's a cell to upload the finished models to Huggingface. I went on there earlier assuming I can resume training on my already pre-trained "x" epoch zip (pth and index file). The training was successful however, I went to go add a repo link to the Huggingface cell and it didn't work. It told me to run Index Training. That also gave me an error. Basically, I wanted to resume training on the models in hopes that it'd sound better on an RVC huggingfacerepo for audio bits.
I do inferencing on one Huggingface repo, thank you, though! :))
Thankyou
Oh, I see. You can only resume training if you have exported the model earlier, which is a huge reason why I hate rvc disconnected. Sure, the cpkt gets saved in the logs, but thats only temporarily because you need to export the model's epoch and step count to create a 2333333 file
Now can you tell me the most beautiful model with the function to separate backing vocals and vocals from vocal-only data?
Ah okay I have those files. Do I zip those with the new pth?
I noticed it updated my original pth files
Im sorry Im asking lol Im new to this
thank u for helpin
However the 2333333 are .pth files
Don't we need index for .zip files to unload into something like gradio?
try this, if not the model can't be trained further https://docs.aihub.wtf/rvc/cloud/rvc-disconnected/#instructions
Last update: Mar 8, 2024
BVE is so bad but thats all we have
Ah so when I trained the model a few days ago, I obtained the .pth file and index file with no issue. However when I resumed training via RVC Disc'd it just updated the .pth file and added those 2333333 files. I'm guessing model cannot be trained further?
yeah you didnt export
so.. pls use something else 
-kaggle
- UVR5 NO UI by Eddy
- How to use RVC Mainline Kaggle by Cauthess
- 🍏 Applio RVC by IA Hispano
- ✨ RVC Mainline by Hina
theres so much better colabs too
I didn't have an account with Huggingface then....lol
......
Alright.
Thanks for the help, though.
Truly appreciated.
np, you just have to read the aihub guides mostly
Mmm, alright, I also like your pfp 🐈⬛


i dont have a cat but that was my psn profile too so it stuck
Cool!
By "
"
-kaggle
- UVR5 NO UI by Eddy
- How to use RVC Mainline Kaggle by Cauthess
- 🍏 Applio RVC by IA Hispano
- ✨ RVC Mainline by Hina
have you tried this?
これを試しましたか?
https://mvsep.com/
BVEは今のところ最高だが完璧ではない
。
。、
。、
。、 やってみます
-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
- UVR5 NO UI for Google Colab, by Eddy Google Colab
- UVR5 UI, by Eddy Google Colab
-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
hello, i wanted to ask if i can stop the training for now and then continue from where i stopped, i'm like 2h24mins, is it safe for me to stop it then use a checkpoint ? how much progress would be lost if i stopped now ? i set it to save every 5 epochs and i'm in the 34th epoch now
do i wait till the 35th epoch is complete ?
yeah 35 epochs. What are you using?
Applio, i'm doing rmvpe
k, it will autosave in your weights folder. The processing and feature extraction is already saved in your /logs/ folder
so you dont need to rerun those twice
you can close out rvc and retrain another time 👍
wait but, how to feed that checkpoint to applio next time ?
Ayo? @opaque lynx level 2 !!! 
your tensorboard will display it all so you dont need checkpoints
basically just delete the epochs that you dont need if you want to save space
oh does it use the logs ?
but they're probably so small anyways. No it's in your assets > weights folder
or am i wrong
I also found something a bit strange that task manager is showing 0% gpu usage
it's in logs > weights but yeah
OH HELL NAH
i got like a 4060 gpu so it was very weird to me
it will be able to resume from the latest saved ckpt
it's like taking less and less time to do epochs though, im getting like an epoch each 2 mins now
I'll try to stop it then change the gpu index, maybe that's the whole issue
if you have your 4060 as the only one, no need to do so
No, i got a dedicated and an integrated one
the integrated one being intel UHD
from my task manager i'm seeing 5% usage from UHD and 0% on 4060
while CPU is 25%
memory usage is almost 100% though
iirc I ever had a pc with intel 14700 (non-F) and 4070, and I didn't need to change that configuration tbh
it's just a bit weird to me what task manager is showing me
OH NEVERMIND i'm seeing 100% peaks

if each epoch doesn't take ages to run, it should be normal
is it normal though that i'm getting an epoch each 2 mins
actually no
last epochs took 8 mins
how is the system ram usage tho?
nevermind that too, i was looking at the time stamp 😭
last one took 2:50
ram usage ?
it's at 82%
anyway training could only be able to run under CUDA, so if it runs, yeah your 4060 was being used anyway
oh true
thanks for the help anyway :))
really appreciate it
that moment when the gpu usage is 0% was probably upon saving a ckpt, or something
actually it's mostly 0% then just peaks at some points
Try checking in gpu-z. Task manager usually shows 3d pipeline utilization which is rarely used in computations
you'd rather want GPU-Z or MSI afterburner to monitor the gpu & other usages
true
i got a bit lazy to use it
and you can set the refresh interval
Ayo? @opaque lynx level 3 !!! 
Can I make a guitar model with rvc and replace guitar tones on instrumentals?
yeah if you have a good guitar stem
I should be able to make one
does using A100 instead of T4 for Google colab make a huge difference in quality or just they both sound the same?
like what’s the difference in them if using them for colab
I’m thinking of js buying google colab pro but I’m not sure honestly so I js wanna know
Ayo? @ember patio level 1 !!! 
Training speed will be faster, but quality won't change
So there’s no real benefits other than that?
can you explain to me like the setting because ion get it. Well advance settings like crossfade all that stuff would it be of use to be in general or is that stuff js best left untouched.
there are some other affordable rental solutions
what’s the best then
vast.ai, AWS, that's what I heard so far
I’m still rly new ngl I’ve js been using this one https://colab.research.google.com/github/hinabl/voice-changer-colab/blob/master/Hina_Modified_Realtime_Voice_Changer_on_Colab.ipynb?authuser=2#scrollTo=lLWQuUd7WW9U
But
They have Google colab?
yeah that's good too
if you have some docker/containerization knowledge to set it up
cuz idk for some reason some links work better then others
😭no clue how
Wait, I though you were asking about rvc and training... But for realtime, the answer is still no. You can possibly achieve lower delay, the quality will still be the same
This has 30 hours of gpu. No colab nonsense https://www.kaggle.com/code/hinabl/public-w-okada-voice-changer
Yeah but I have like a bad compile
Ayo? @ember patio level 2 !!! 
oof I thought it's the training notebook
its all on the cloud
So it’s just more efficient than that colab stuff in general?
Higher quality and faster rates????
yes even though its owned by google too. ppl are saying its kinda glitchy though but hina just imported it to kaggle
brooo what 😭
importing models and everything is the same old process
easy right
or is it different
how does importing models work or js are u unable too
they don’t rly say it on the website or
Same as before. Hugging space or pth file upload
You don't necessarily need the index. Also #🔍│help-w-okada
thank you sm
I'd like to get some recommendations for RVC software.
Or you can also use colab or kaggle for modified version 
https://colab.research.google.com/github/deiteris/voice-changer/blob/master-custom/Colab_RealtimeVoiceChanger.ipynb
https://www.kaggle.com/code/suneku/voice-changer-public
Table Of Contents Introduction (with link) Model Loader (Download & Upload) Inference (use RVC AI Voice Models) Ilaria TTS Settings (Inference) Introduction (with link) Ilaria RVC Zero, is an RVC (Retrieval-based Voice Conversion) Fork made by Ilaria & mikus, running only on Hugging Face Spac...
Ooh, use the modified version instead then
It might work better
アップロードに失敗しましたか?
Pthのファイルとindexのファイルのサイズが大きすぎてアップロードができないと言われました
I was told that the pth file and the index file are too large to upload
既存の URL を使用した場合、機能しますか?
試してみるためのサンプルURLはこちらです
https://huggingface.co/listra92/MyRVC/resolve/main/Pekora_e241_s4820.zip?download=true
モデルアップロード機能を使用していないのでわかりません
is TITAN pre-model good ?
Both Index and pth files are now showing red text notation.
you can try taking another model's index file which is smaller, maybe from this example model
このサンプルモデルなど、もっと小さい別のモデルのindexファイルを取得してみるのもよいでしょう。
https://huggingface.co/listra92/MyRVC/resolve/main/Pekora_e241_s4820.zip?download=true
if you don't know much things, the default pretrain is a fine bet, though you can try TITAN as well
横向き画面で pth および index ファイル名を表示できますか?
試してみたが成功するはずだ
それで、エラーメッセージには何と書いてありましたか?
エラーメッセージが出ていませんこの表記だけです
赤文字じゃないですか?
成功した場合、「Status」メッセージに「Uploaded」と表示されます。次に、「Inference」タブに移動し、「Refresh Models」をクリックし、Modelのドロップダウンをクリックして下にスクロールすると、最近アップロードされたモデルが見つかります。
if it succeed, the "Status" message should say "Uploaded", and then go to "Inference" tab, click "Refresh Models", click the Model dropdown, scroll it down, and you should find the recently uploaded model.
どうですかね?
Ayo? @final sentinel level 5 !!! 
ここでご覧ください
Good morning. I want to make a voice model, does anybody know where to start?
-gui
-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
- UVR5 NO UI for Google Colab, by Eddy Google Colab
- UVR5 UI, by Eddy Google Colab
Who do you want to make a model of?
Your own voice or any character?
myself lol
Ayo? @dense hamlet level 1 !!! 
Alright, you can start recording at least 15 or 20 mins of your voice.
My gpu is 4070, so I don't mind doing it on my pc locally. I have a youtube video already of me talking for a while
Should I just download applio from hugging face
and install it
is that the best way?
デスクトップモードでもう一度アップロードしてみてはいかがでしょうか?
I also want to do TTS after I train my own voice
How do you train your models on a long dataset really fast?
Ayo? @summer estuary level 2 !!! 
From where you wanna download applio?
Hugging face right?
Show a link.
You can't
Well, you can't unless you put a high batch size on training.
Training speed will also probably depend on your PC specs unless you use colab or kaggle.
I want to train 1000 epochs model what to do
Your dataset is clean?
Even if it's clean, i wouldn't recommend you to train it to 1K epochs.
My dataset is very clean
depend on the tensorboard graph
In that case, you shouldn't just train it to 1000 epochs.
My dataset currently has 25 minutes
This is also another link to download newest version
https://github.com/IAHispano/Applio/releases
Go for 500 epochs instead. But make sure to check tensorboard
Download that one
出来ましたありがとうございます!!
Thank you
You're welcome.
But i want 1000 epochs what to do
Just check tensorboard.
You never know when your voice can overtrain after or even before it reaches 1K epochs.
Also, more epochs doesn't mean more quality
Have you ever trained 1000 epoch and where is it difficult
Nope. It's not mandatory to train a model to 1000 epochs. Just go for 500 epochs and check tensorboard
ye go 500 first
どういたしまして。
Heya guys, do I actually need to put save frequency to 1 to get the epoch I want, or are there other ways you get it?
I’m using RVC v2 Disconnected in Google Colab.
the ckpt saving process is slow as hell there, btw you can try this kaggle notebook https://rentry.co/RVC-Mainline-Kaggle
This guide for Mainline Kaggle is an alternative option to the Mainline Colab notebook for training voice models
It is complete and should walk you through every step of the way since Kaggle has a difficult learning curve. However, it will be updated constantly since new updates to the notebook w...
-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
-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
- UVR5 NO UI for Google Colab, by Eddy Google Colab
- UVR5 UI, by Eddy Google Colab
ごめんなさいコンバートの下にある設定の使い方を教えて欲しいです
Sorry, I need to know how to use the settings under convert.
-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
-help
-local
- 🍏 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.
Yep, you must put it to 1 or 5 if you want.
Just put "save latest ckpt" to yes
@odd shale can you help me?
The feature extraction cell ran successfully but this is the problem with index traning
I'm not really sure..
@low shard Can you help him?
Ok
Let me run the entire colab again and see if it happens again
And I will give a ping if it happens
通常、次の設定を調整するだけで済みます。
Pitch level: 男性の声を女性の声に入力する場合は +12 に設定し、女性の声を男性の声に入力する場合は -12 に設定します。
ほとんどの場合、他の設定を変更する必要はありません。
-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
- UVR5 NO UI for Google Colab, by Eddy Google Colab
- UVR5 UI, by Eddy Google Colab
Meaning it ran fine
Nope, if it says no-feature-todo it means it didn't work
You don't have to ping staffers everytime.
how are the zip contents?
The dataset has 10 songs on it so....
Why
spaces in the name 💀
You should not put spaces on your dataset.
Ohhhhh
Ayo? @calm jackal level 4 !!! 
note that if there's 0's in the preprocess that means nothing was processed
Ayo? @knotty moth level 23 !!! 
Is this fine
Ok
Is the name fine?
You can use a shorter name if you want.
Ok
-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
- UVR5 NO UI for Google Colab, by Eddy Google Colab
- UVR5 UI, by Eddy Google Colab
anyone know how to continue training my own model?
I trained my model to 400 epochs then i restarted my computer and tested. Now i want to continue training it to 1000 epochs but idk how to continue
Im runing RVC locally
Ayo? @lucid charm level 1 !!! 
just same settings as before
We got a tragedy making RVC models before gta 6
Last night I was trying to train my model on the RVC GUI from the Kaggle link. But midway through last night it gave me a connection error and now I can't even visit the site. What did I do wrong and what should I do to fix it? :(
(normally) if it's the rvc link, you gotta wait till it finishes loading
Oh okay. I was just wondering if it had somethign to do with last night because the connection timed out for some reason
Wait until the cell finishes loading?
Because it like goes to the website for a split second like normal and then randomly is like nope you can't use this because the connection won't let you
I mean after the console shows kinda "found a GPU T4"
have you tried using a correct ngrok token, or even using a vpn?
It's working now for some reason. I have no idea what its issue was. But I know I used the correct ngrok token
Sorry about that 
Hello
I want to train the model, but I don't have a PC, so is there any way to train the model?
Ayo? @cerulean moat level 1 !!! 
rvc disconnected
are you using ngrok now?
How long should the dataset be used to get the best model?
probably try again next time in a new session 
i personally do 15 minutes for the best quality
what is the best f0 change for male to female singer? 12 sounds not natural from the original
12
12 wasn't good at all
wdym
Ty
Yaaaaa I decided to use the RVC Mainline because somehow it was easier to understand than the other one
uhh its like pitching up like chipmunk
mainlien the goated
then use 0
HAHA :D
what softwares have rvc v2?
rvc v2
Ilaira RVC, Original RVC WebUI, Applio, Mangio, etc
Sorry wasn't much active today, is the issue solved?
AHD didn't talk anymore, so i guess so
and all of these work like voice changers? using vb-cable?
Oh for real-time use w-okadas
?
You know it's supossed to say "I want to hold your hand" buuuut it generated weirdly
can someone help, my voice sounds super robotic and glitchy
problems with the server link on google colab error 404
Ayo? @rain urchin level 1 !!! 
Which colab link are you trying to use?
no, the google colab link
idk what server ur talking about
be sure you arent watching youtube videos
I think you should be using https://docs.google.com/document/d/e/2PACX-1vTIceEcBfS6Zqolv_QEysrFfI_EJikPxozWptP_EjkpLVl-l-gdo-ijBonQMTviAHEYm5emmd9k9TdC/pub instead
thank u very much
yw
Thing is I can't seem to get RVC V2 models to work with it. Only rvc v1 or whatever 😕
RVC v2 models should work
which models did and did not work?
Literally only the base models worked for me. Idk if I downloaded the wrong version or whatever. Could you dm me the correct link? I downloaded from the github but still, need to know if its something wrong with my pc or if I just failed w downloading the correct version
Ayo? @whole cypress level 1 !!! 
What’s your GPU
The base models were onnx so that might be your GPU
Nvidia 1660 super im p sure
This is a really weird issue but I've tried easyGUI, Illaria and Mainline Colab and when I load a .pth file into the weights folder/go to use a model I keep getting this
Use Ilaira RVC on hf for inference instead of colab
Does HuggingFace have batch rendering?
Not Ilaria RVC
RVC ZERO might but I’m not sure
-spaces
- 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
Make sure the spaces your using say running on zero
Otherwise conversions will be painfully slow
Nope RVC zero doesn’t have batch
Applio does I'll see if that one works 😮
I'm working on an RVC WEBUI. What's the appropriate amount for Epoch? Does the more you upload it, the more the sound quality is different?
I had the same problem
u fixed it?
what is the llvmlite.dll error
it says its missing
i can't find it in the packages
have you tried another version which is either the latest 18a or this fork? https://github.com/deiteris/voice-changer
||btw why are you flagged as "spammer"? was your account hacked before? ||
@wanton iron, I have found 5 results that match your search!
@wanton iron, I have found 1 results that match your search!
Sorry, but I could not find models that match your search "Jay Cortez"
Sorry, but I could not find models that match your search "Jhay Cortez"
@wanton iron, I have found 7 results that match your search!
@wanton iron, I have found 7 results that match your search!
@wanton iron, I have found 3 results that match your search!
@wanton iron, I have found 1 results that match your search!
nah lmao
bro idk where
nope, go to the releases section https://github.com/deiteris/voice-changer/releases/tag/b2241
and i want colab one btw @knotty moth
cause my device is shit intel
i want a software can handle it fast
Either use this
https://colab.research.google.com/github/hinabl/voice-changer-colab/blob/master/Hina_Modified_Realtime_Voice_Changer_on_Colab.ipynb
Or mod version
https://colab.research.google.com/github/deiteris/voice-changer/blob/master-custom/Colab_RealtimeVoiceChanger.ipynb
you just need to run it in your browser, no need to download anything and open cmd shit like you had done before
ah okie
@knotty moth sorry for bother the site wont work 🤓
hahahaha
hahahahaha
my situation xdddd
how do i make an ai cover
Check pinned message for Ilaria RVC Zero + Guide
what download should i use i have a Nvidia GPU and Intel CPU
What program are you trying to install
RVC voice changer
theee
w-okada voice-changer
unless there smthn better
but i think i downloaded da wrong version for my pc cuz its bugging a lot
You can either install the normal wokada, follow nvidia steps + virtual cable VAC
https://rentry.co/VoiceChangerGuide
or you can install a modified version. It may run better than the normal if you dont know how to set up a venv, so id go with the 2nd. follow nvidia here too + virtual cable
Maybe its important to know what nvidia gpu you have?
if you could tell me
rtx3050
Ayo? @mossy estuary level 1 !!! 
why does MMVCServerSIO_win_onnxgpu-cuda_v.1.5.3.18a Run Poorly but Dubbing Ai works so good?
do they both not run off my pc?
how can i make MMVCServerSIO_win_onnxgpu-cuda_v.1.5.3.18a better
how do u clone your own voice
can i manually download both w-okada and deiteris modified version? i want to test the difference
Ayo? @rose arrow level 2 !!! 
im concerned if it will cause problems or not
As long as you keep and run them separately, there's no problem
おい
thanks
I would reccomend @wispy lodge's advice, but next time you ask for realtime voice changer advice, make sure you ask in related channels like #🔍│help-w-okada
how do u clone ur own voice
Hey there! let me pull up the guide for you.
should be
👍
Now, voice cloning takes some time and might be hard (depending on your computer expertice) to do, so you can either:
A: do it yourself (https://docs.aihub.wtf/essentials/how-to-make-voice-models/)
B:Hire someone like a Model master or a Helper or anyone else with experience
In the context of RVC, the dataset is an audio file containing the voice the model will replicate. It can be either speaking or singing.
Wish you good luck in your RVC adventures!
what should i do in recording vocals
idk what to say for 10 minutes
im trying to make an ai cover so should i just sing
find a book to read
okay
dont sing, it might lessen your range
okay thanks
it picks up all sounds, make sure to enable sup2 and move s.threshold to the right
try deiteris fork
What’s that
@flint geyser
ok i got rvc web ui now what
i also made the voice models now how do create the model
- 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
- UVR5 NO UI for Google Colab, by Eddy Google Colab
- UVR5 UI, by Eddy Google Colab
I created a model and am using it, but when I first speak, the volume is very low, and after about 0.3 seconds, the microphone volume returns to normal. I thought it was a microphone problem, but other paid models do not have this problem. Is there any way to improve this?
ciao
ciao
can you explain better?
è possibile creare il continuo di una canzone se vi mando un pezzo di canzone
Ayo? @kindred tinsel level 1 !!! 
Literally, when I use the model I made, the sound will be low when I speak for the first time, and then the microphone will function normally after about 0.3 seconds. I thought it was a problem with my microphone, but the paid models are working normally. I'm curious about what the problem is now.
The sound is very low when I first start talking, and then the microphone volume returns to normal...
puoi farlo con suno
okada settings?
se ti mando la traccia potresti farlo tu?
no, non posso.
è semplice da fare, basta uploadare la canzone e cliccare su estendi traccia
I don't have these symptoms when I use someone else's model. Did I make the wrong model?
maybe(?)
everything depends on the dataset mostly
in this case at least
did it have bad EQ?
Ayo? @steep elk level 11 !!! 
The sound quality is a very good model.
did you use any pretrain?
what were the settings
try targeting 500 epochs then analyze the tensorboard graph
@proven hill@knotty mothThank you for your help👍
dataset quality is more important, epochs here are probably not the issue
thx!
-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
- UVR5 NO UI for Google Colab, by Eddy Google Colab
- UVR5 UI, by Eddy Google Colab
Hi everyone, sorry but now that the colab site has been closed, how do you do it?
türkçe bilen anlayan birisi var mı? Program çıktı verirken hata veriyorda güncel program var mı?
what you need to do?
which colab and for what purpose?
Covers or training?
ikr
there used to be a RVC model of you but it got deleted 😔
so do i have to put the voice i want to change in again?
ye
so sad 😭
alright thanks. Time to waste another 50 minutes waiting.
the google colab site itself is not closed, probably a colab you are talking about is broken, what do you need? Inference (use the model) or train (make the model)?
could anyone help me w this
What are you trying to do?
Try installing the pre compiled version
can you send me a link
Hi guys
Can anyone help me understand why my ai changed voice sounds bad
Its so mumble
Send a screenshot
those versions are outdated tho
also this is what i get after running applio
any help please🙏
Yo im using ur rvc how long will it take for 300 epoch
i have 3070 ti
rtx
idk you do the math
My real time gui won't load, do I need to redownload it?
ok i did it how do i download them now
the model and index
@proven hill
please dont ping me im not an helper
sorry
you find them in the logs folder (index) and the weights folder (pth)
theres nothing in weights
Ayo? @mystic tangle level 5 !!! 
It should have it. did you clone repo or download rvc1006
I'm having trouble installing the latest local version of RVC because of this
I've installed from their official website several times and the zipped folder never contains the go web.bat file
Ayo? @real field level 1 !!! 
Did you unzip it
my model finished training but theres nothing in weights or logs
omg ty
Doesn’t include it for some reason, you can get the file right or the repo
this?
Yes
just put it in the Intel RVC folder and run ?
so what now
Yes
Open the install bat
okay, ty❤️
why not use orginal RVC instead of applio
Wait
Are you sure you downloaded the precompiled version
well im not sure 
are you sure its finished?
mainline rvc the goat
u have to install Microsoft Visual C++ Build Tools:
1- Go to the Visual Studio downloads page, here is the link:
https://visualstudio.microsoft.com/visual-cpp-build-tools
2- Download and install the "Build Tools for Visual Studio".
3- During installation, under "Workloads" (or individual components), select "C++ build tools" and ensure the latest versions of "MSVCv142 - VS 2019 C++ x64/x86 build tools" and "Windows 10 SDK" are selected.
4- Complete the installation.
After this, try install Applio deppendencies again
omg sorry
because its 3.2.1
the og rvc is older im pretty sure
LOL
but better
i love you bro
Yeah but applio doesn’t have anything special
try 3.1.1 I suppose? 
mainline is still faster
try? to me?
idk if u remember me
try it
i deleted my disc
i will thank you bro
@devout field use mainline instead of applio
i dont member :(
sadge
sorry nothing personal i take meds for it im so sorry


AI HUB Docs

