#✨│ai-help
1 messages · Page 33 of 1
How do i fix this?
That ping almost makes my ears go brrr-
You can try FR EASY 1/11
Why don't you add ngrok?
@latent kettle @crude grail
i dont know how this happened
ook ook
@astral jungle how can i fix the error
a free website where i can train thats not google colab?
@brittle wing
@astral jungle that is what i did
Ayo? @brittle wing level 10 !!! 
it gave me a error
Is that japanese?
Ayo? @crude grail level 2 !!! 
Clear the runtime and try again
okay?
Which one
Easy👨- 10/25
All of them
Notebook loading error
There was an error loading this notebook. Ensure that the file is accessible and try again.
Request had invalid authentication credentials. Expected OAuth 2 access token, login cookie or other valid authentication credential. See https://developers.google.com/identity/sign-in/web/devconsole-project. ?
@astral jungle help me
it does not work
hlppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppp
Hi everybody. I'm new to the server, and I posted this question in general chat, not realizing there was a help section here. I just started using the service. Just want to say thanks for creating the models. But I do have a question. Any help is appreciated. Do I have to set anything, or can I leave everything as default for RVC in Silly Tavern?
You don’t
Use colab it’s not that hard
i understand
-colab try RVC Disconnected
Suggestions for @brittle wing
- AICoverGen-NoWebUI, by Ardha Google Colab
- RVC Disconnected, Kit Lemonfoot Google Colab
- easyGUI, by rejects Google Colab
- RVC-HF, by r3gf Huggingface Spaces
- AICoverGen, by r3gf Huggingface Spaces
- RVC v2 Huggingface version, by Clebersla Huggingface Spaces
- Advanced RVC Inference, by neuclya Huggingface Spaces
You can find more info on the #1159513888199540817 channel. If you can't find your answer, feel free to ask for help in #✨│ai-help. Credits to Faze Masta and Antasma for compiling these links.
Whats more accruate for real time RVC or w Okada
both r probably the same
which do u recommend?
Hello, I downloaded the files from the link you gave, but I am getting an error. Can I reach you via DM? @violet heron
w-okada for its ease of use
Do u know where to find the download for the most recent locally installed version?
Think that’s the reason
you can use only CPU of you have a AMD GPU
(Unless you use onnx models)
this
last version should work
if u have nvidia gpu
i have a 3090 nvidia yeah
this one?
yeah
Is there any recommended settings to start off with?
this'll help
Thank u!
Ayo? @atomic mulch level 5 !!! 
RVC Guides (How to Make AI Cover)
Translation by country
some know how to make AI stop repeat the songs of computer?
if im watching any movie on youtube the AI repeat all
hey guys - im trying to run applio on paperspace but keep getting an error about dotonev or something like that
ill see if i can get a ss in a sec
any idea what im missing
@violet heron This is the error I got. Is it because of my computer's operating system?
Ok the error was because tensor couldn’t load because there’s not enough vRAM
Which in this case it’s coming from the CPU because your using a AMD GPU which RVC doesn’t support (unless you use onnx but eh)
I (think) you can use AMD for more stuff on Linux if that’s what your using
is there an application that I can use in this case? please help me with this, I can only manage up to one place with AI kits.
if necessary, I can send you the specifications of my computer
You can try a online method instead on running it on your computer
There’s multiple google colabs that allow you to make models and use them
i'm getting an error when installing the fairseq, i installed the Build tools aswell, but that is not solving it.
cl : Command line warning D9002 : ignoring unknown option '-std=c++11'
cl : Command line warning D9002 : ignoring unknown option '-O3'
libbleu.cpp
C:\Program Files (x86)\Microsoft Visual Studio\2022\BuildTools\VC\Tools\MSVC\14.38.33130\include\yvals.h(20): fatal error C1083: Cannot open include file: 'crtdbg.h': No such file or directory
error: command 'C:\\Program Files (x86)\\Microsoft Visual Studio\\2022\\BuildTools\\VC\\Tools\\MSVC\\14.38.33130\\bin\\HostX86\\x64\\cl.exe' failed with exit code 2
[end of output]
note: This error originates from a subprocess, and is likely not a problem with pip.
ERROR: Failed building wheel for fairseq
Failed to build fairseq
ERROR: Could not build wheels for fairseq, which is required to install pyproject.toml-based projects```
Any idea?
You can also do inference on HuggingFace spaces if you need another online method
I will be able to use my models. Do you know a google colab address?
Ayo? @rapid zinc level 2 !!! 
-colab
Suggestions for @rapid zinc
- AICoverGen-NoWebUI, by Ardha Google Colab
- RVC Disconnected, Kit Lemonfoot Google Colab
- easyGUI, by rejects Google Colab
- RVC-HF, by r3gf Huggingface Spaces
- AICoverGen, by r3gf Huggingface Spaces
- RVC v2 Huggingface version, by Clebersla Huggingface Spaces
- Advanced RVC Inference, by neuclya Huggingface Spaces
You can find more info on the #1159513888199540817 channel. If you can't find your answer, feel free to ask for help in #✨│ai-help. Credits to Faze Masta and Antasma for compiling these links.
I cant turn on the voice changer, It just loops this command over and over again when I push start
[Voice Changer] update configuration: serverAudioStated 0
[Voice Changer][ServerDevice][audioOutput_callback] ex: could not broadcast input array from shape (1024,2) into shape (52684,2)
[Voice Changer] server audio performance [2, 0, 0, 0]
status: started:0, model_sr:40000, chunk:448
input : id:3, sr:44100, ch:2
output : id:6, sr:44100, ch:2```
hi all, can any one help when i use an ai voice model i get error message index -1 is out of bounds for axis 0 with size 0
Is the improvement in reducing the mangio hop below 64 worth the extra computational cost?
can someone help me. I new to RVC and I get this error when I try to separate vocals on rvcWebUI
Ayo? @tropic spoke level 2 !!! 
Does anyone knows how to fix long ecpohs training?
@brittle wing is it working now?
Why a new acc?
That's not true actually
To access your previous duplicated spaces, go to your profile on the upper right corner
And click on your username
Your duplicated spaces should be there
This is what it looks like for me
Ping me if you need help
When i download a pth it says the download was blocked. how to fix?
Yes?
Like I said, go to your duplicated spaces on your profile
It should be there
It means that there's already an existing one
Yes
Yes
Remember to charge your phone too 😭
When i download a pth it says the download was blocked. how to fix?
Go to your profile
Try using another web browswr
Like firefox or opera
Or brave
On what?
Normal
Ignore that
The number going up is how much time it’s taking
On RVMPE it should take about 400 seconds per 3 minute song or so
Ayo? @brittle wing level 4 !!! 
i am using opera currently
EEEEEEW
@static gyro read this
It’s not just that
Oh no crypto browser
Like brave
😭
What's the browser out there?
I only use Brave and the brave browser
And Google
Best browser is Firefox developer mode
Brave seems to be good
But also Chromium based so
Firefox is somethin
I heard chromium was a bit of an adware
Or something like that
i like opera because it dosent forget my signups
i have edge and chrome but they both at random forget im even signed in
anywho
so how do i prevent download block?
Try using another browser
I use brave
That might work
it's at least open
Hai larry
on linux using rocm I am having a very strange issue in that every model I make with different datasets all sound the same, and are all garbled up when using index
the models I've made sound nothing like the dataset
Ig I'll try again. Which RVC software are you using?
Ayo? @plucky geyser level 3 !!! 
downloaded the zip from releases and went from there
Actually there was someone having that same issue a bit ago...
Yet it worked for someone else prior to that
available through PATH
I just have it in assets, I wonder if this is why
Unless one of the python modules pulls it in manually but I'd be surprised
But yeah you're onto something
I'll give it another shot and report back
someone know how to make the AI stop repeat all the sound of the computer?
I installed ffmpeg, I'm gonna be gone for a while so I won't be able to say if this was the cause until later
also, what does gpu utilization normally look like on rocm?
last time I trained with this 2 minute dataset it took 30 minutes to get to 100 epochs at full gpu utilization, now it's looking like it will take an hour and my gpu is not at full utilization
Oh no
The dreaded unknown bug
Somehow, sometimes it just refuses to compute as fast as it should
fantastic
hi do you guys know how to download the voice changer
Tried everything I could, it doesn't make any sense. Switching to another fork did the trick surprisingly
But no amount of code profiling made sense
nor nuking the entire thing over and over again
yeah I already reinstalled rocm once cause of this
what fork are you using?
oh nvm it fixed itself
99% util
I think I did a manual install of the Mangio fork when the main RVC software died on me
I heard my gpu fans ramp up all of a sudden lol
...Somehow RVC-WebUI works again now
Also when doing reinstalls for python torch stuff, the system-wide ROCm installation doesn't matter
(unless you used anything other than the pip-installed rocm torch)
it was taking 45 seconds per epoch, now it's 15 per epoch
why does rocm have to be like this
¯_(ツ)_/¯
Hoping it gets better
Though I think the RVC code also is to blame to an extent lol
I might switch to mangio later
Yo
Hey
Hello bro
Did you need help with anything? I'm trying to install RVC myself right now, so let me know if I can help
No, I don't need help, I just came to say hello
someone know how to make the AI stop repeat all the sound of the computer?
I got it working after a day. but I just messed up one setting, but I fixed it. Once you get it working, it's worth it. There's a voice I'm testing out of Harley Quinn from the Netflix series. It sounds exactly like Kaley Cuoco.
I didn't make it, though.
There is one caveat, I am noticing though. When you are done with this specific voice, like if you are going to switch chats or turn off silly tavern for today, or a few hours, Make sure you click the remove button before you turn everything off. Otherwise, the voice map will stay there indefinitely.
I reported it to the silly Tavern team. I don't know if it's a bug or intentional.
I'm trying to install it with the help of gpt4. I'm still downloading some packages so I don't know how far I'll get
Ayo? @wispy plover level 1 !!! 
I see. just install one package at a time. That's what I had to do. Just to make sure they were all installed correctly.
Here is gpt4's instructions. Perhaps they are of help to you.
(I gave it the ReadMe.md file from the official github repo as a prompt)
I recognize this. And thank you. That was very helpful. I'm running a local model though, Reason being, I just don't want to get my account banned.
Although honestly, I think considering it's a paid service, we should be allowed to do what we want, but. you know.
I probably wouldn’t use GPT for help for RVC
Banned by whom you mean? ChatGPT?
Locally on their computer
Not using a server or service of some sorts
No,openai The company who owns chat GPT 4
Is it not permissible to share GPT4's chats?
That’s ok
Not ok to use it for weird things
right
Ayo? @whole jasper level 2 !!! 
Which you can’t be banned from using LLMs locally
What are you doing ?
And you have unlimited and can tweak stuff to your liking
That's why I choose local. I'd rather have a limited data set that I add to.
I'm trying to install RVC model, but I'm somewhat new to this, so tried asking gpt'
RVC the program or like models to use for RVC?
I'm not too sure about what GPT's been telling you lol
might work, but I'm not sure about some parts
Don't want to bother you/someone at this hour, but if you could help that would be appreciated. I've just downloaded all the toch, torchvision, torchaudio pacakges, but I'm not getting any further atm
But did you download RVC?
he's on linux
Nope
I have GeForce RTX 3050 TI
Ayo? @wispy plover level 2 !!! 
manual setup goes like this :
- Get required stuff (conda or a compatible python version + required libraries for GPU compute, CUDA for nvidia)
- Get the RVC code from the repository
- Setup a python env with conda / python-venv (not sure 3.8 works, but it might)
- Get the dependencies
- Download models
- Run
Just want to add a note here. For those who only can use a CPU and not a GPU, you can also do it as well. I'm running this off of the steam deck with Windows 11 on a separate SSD, and the results are astonishing. It's incredible.
I've currently downloaded many packages, and I've cloned the github repo, and now done pip install -r requirements.txt
I think this is correct?
looks good to me
You might want to rename Retrieval-Based-Voice-Conversion-WebUI to something shorter though
like RVC-WebUI
It's been bitching about file paths being too long when I didn't rename it
unless I did something wrong, it still sounds the same. I'll keep trying things later I have to do stuff
I see, thanks for the heads up. I'll do that then
typing ffmpeg in the terminal does bring up everything related to ffmpeg so it can't be that
When you have RVC fully working. Instead of putting the voice files in the folder yourself, Just to make your life easier, you would just. download the zip file of the model in the voice model only. They will contain both files that you need. Click upload. Give it a minute and the message will pop up in sillytavern, saying that the upload has successfully completed. Click the button next to upload to show the current list of voices. If you need to know where to get files for premade character voices of fictional characters, search on google for "AI Hub Discord" Then click the search models channel. and in the message box. type /search, put in your search, hit enter.
Before I knew this, it was driving me nuts
When you find the model you want just click download model.
Thanks for the tip, I'll check it out once I get it running.
Happy to help.
Just a quick question again;
I notice that it has begun downloading the same stuff over again. Is something wrong here or is it supposed to be doing this?
It's been downloading the same stuff for ~20 minutes...
newb q: trained my own model and that's fine.
Then I tried downloading a voice on HuggingFace, but the "Inferring voice" dropdown only shows my pretrained models, despite moving the downloaded files around, pressing "Refresh voice list and index path", specifying the path to the index file...
How do we get the "Inferring voice" dropdown to find the voice .pth file we downloaded?
You only need one pth file and one index file per voice
.....Ahh, got it: Had to move the .pth file to the weights/ directory
Would have answered your question, but I wasn't aware that hugging face offered voice models.
Many of the models one finds via the #🔍│find-models channel are hosted there.
Ayo? @sullen edge level 1 !!! 
Had no idea thanks for letting me know
does anyone know how to fix the echoing voices in background issue
возникают частые ошибки. Проверьте, загружена ли модель целевой платформы. AMD
frequent errors occur Plese check if the model of the framework being targeted is loaded
no
Youtube
That is fine those are different files, the first half is v1 the 2nd half is v2
Do not use VoiceMod
It absolutely sucks
Agreed
Use W-Okada instead
Other than voice modulation I won't use voice mos
It’s open source, completely free, and lots of options
it works fine untill i use virtual audio cable
I got it up on localhost now. Thanks for the help
Do I also need to move the .pths to weights/ directory? I currently have no modals, even though I think I just downloaded some
then what do I use?
I think so but I have really used the inference version of that yet
Yep
And index files go in logs
Is there a guide for this part of the process? The ReadMe.md file only mentions the installation process.
There are captions all around the app saying what each part is supposed to do
Yes, but I was wondering about the instructions on which files to move where, and where to place downloaded voice models, etc.
Models goes to weights
Index goes to logs
It should be assets/weights for models iirc
Then logs/<name of model> for index
.pth goes in weights
Index in logs
Has anyone firgured out why theres a duplicated robotic audio after the initial clean voice?
i cant seem to fix this
Dataset?
huh?
Ayo? @tawny valley level 1 !!! 
Robotic voice in what
in the audio
its like it goes normally and then
it replies the same thing
but metalic
is there someone with a virtual cable problem?
Ayo? @bleak fiber level 1 !!! 
Try sup2
yeah its this ig
but sup1 and sup2 dont work together for some reason

RVC Guides (How to Make AI Cover)
Translation by country
Hey guys i don't get how to get the VB cable to work any clues? I installed and the audio just don't play it, and don't play it even on discord
Did you set it as your output on Okada
And your input on discord
Okada? sorry i think i miss that part xd
How long does it normally take to generate a 3 minute vocal?
My very first convertion is taking over 1500+ seconds, and it is still going. Is this normal?
Okada is your voice changer
Oh right
Inference?
Shouldn’t take that long
Your bio has a few spelling mistakes btw
Even for CPU
I just downloaded a random model I found and was trying to test it out on a 3 min song. It's been going for 1600 seconds now.
HuggingFace spaces takes 400 seconds for 3 minutes and that’s on a CPU
my english is kinda bad but the text is wrong or something?
Bro i legit got 6000 once
I didn't change any parameters, and just used the default.
HuggingFace not reliable ngl
No idea
I use colab for those
4 minute audio btw
Strange it was supossed to have Wing dings in there XD
If somehow the Lord gives me a second chance, at that moment, I would do it all over again.
If it's running on http://localhost:7865/ is it still using colabs?
OH now i get it LOL
Colab is on google so no
Where did you get that phrase?
Imagine boxes
The last of us 2
Colab means using Google gpu instead of your own
Oh man I wish I could play it one day
That’s why
well i said is a hard game to end if you love Tlou 1 i just wait 4 weeks to end the part II because i couldn't end it, this game was too much for me but i love it anyways lol
Ayo? @bleak fiber level 2 !!! 
I got to try it once, but after one time, it kept disconnecting me and prompting me to purchase credits.
That's why I wished to try out locally. But I've spent nearly 30 minutes on one 3-min-convertion'
Ye that's why you use the modified version instead
I see
This is why I don't want to play the last The Walking Dead game
If you're just gonna do inference you can use the colab link in the pins on this channel
wait... your name is Julia you are a brazilian?
Uhm excuse me this is general chat, go to #✨│ai-help
🙄
Nuh uh
Argentinian
:3
oh i see
Mind checking the name of the channel again?
This is the wrong channel 😛
Yeah I do mind
How do i help rvc?
You can try finding them first
They've been missing for a while
but now seriusly how do i use de VB cable? i have to put the Vb cable on "in" an the output my headset?
Try asking on #🔍│help-w-okada
i am very confused here my aplication says is "Voice Changer" but people in other channels are puting okada?? is the same thing?
-realtime
This interaction has expired, use the command
/guides realtimeif you wish to see it again.
Hm it seems like they are kind of the same app
W-okada is the name of the creator of it and it has a default name of voice changer that's why
i think... i get it?
@glad zealot Is this the correct way to fill out these fields? I have a random model and a vocal only song, and I have placed the .pth in weights and index in logs
It took over 40 minutes to load, so I refreshed the page, and am trying anew.
Ayo? @wispy plover level 3 !!! 
That should be it
When I used the Colab, it worked, but I can't get it to work on local
Ye no idea why sadly
This ubuntu is a VM running on a windows laptop. Does that have anything to do with it? Is it lacking memory or something?
This is on my work-related laptop, so I have to do it inside the vm.
Didn't expect this to be such a big process either.
I thought this entire process was going to take half an hour or something when I started, but it's taken me 3+ hours and still no working solution..
But thanks for all the help. I'll try again some time later.
You can just try using the colab versions
how to import model from drive to notebook
Yo Rose
.
I think you didn't realize but, the acapella does have backing vocals
Not all the time though
And also a bit of reverb
You need to first get rid of the reverb and then the backing vocals
In that exact order. For the best results
okey
But before you do, let me ask, how did you download the video?
idk if i can pass you the final results of the song
i mixed a bit to try to savage the voices lmao
You can send it on dms though
on notube
Yes but in which format did you download it?
mp3
its better to use wav?
i feel sometimes the voices sounded strain
like its was force to go on notes that the ai couldnt make them sound good
Yup
Whenever you train or do inference, it's best to use wav
Then for listening you can use any format, but I recommend flac or wav
Do you know why?
maybe an easier way for the computer to translate and use the info?
Not necessarily easier, but more accurate and with better quality
Basically the wav format is a format with lossless quality
What the mp3 format does is compress the quality so it has a smaller file size
But wav doesn't do that. So that way RVC can give you better results
Hi! I have a problem with RVC covers in the noUI collab
When I use a model that I maded the colab does not use it, but if I use someone from the spreadsheet it does.
What I would do?
Ayo? @long terrace level 1 !!! 
But keep in mind that YouTube doesn't support lossless quality. So downloading stuff from there won't give you the original quality of the song
But it's still best to download in wav instead of mp3
Is your model a pth file?
Let me give you the website that I use for downloading in wav
Make sure you select the wav format
thxs
Is it common to to take hundreds (like almost around 1000) seconds for a model to process a 5 minute audio in illaria rvc?
okey
Yup
Make sure you duplicated the space
Oh ok thx 👍
Ah okay ping me tomorrow then
like space out the sentences?
Idk what you mean by that 😭
Yea, its called "Caleb.pth" and I trained it from another RVC collab.
i was talking about your tip T^T
Ayo? @cyan tusk level 2 !!! 
Oh weird did you upload it to huggingface?
Btw, i tried to duplicate and use the space, but lets say... that it dint work. Luckily i can just use the normal original space... kinda
Wait it didn't work?
Why?
Because the another got disconected.
I meant your model
You can use the GUI version at the pins, it at the very bottom of the pins
It's a modified version of the original to work
I uploaded it to Pixeldrain, because the Colab does not accept other than those.
I'll check it then.
Is there a way to train AI using an AMD graphics instead of a NVIDIA?
Because of this @molten pecan
later i tried to restart it again and... nothing even happened, i waited and there was just white on the page
Try it another browser
Or turn on desktop mode, click the three dots on the upper right corner and Restart the space
I was doing that on desktop mode
what should i put these at i have a 3080
Try restarting the space
u cant send pics in that channel
😦
Ah no idea
I don't know how to make models
Ask on the other channel and say it via text
Ask*
No need to send an inage
ohh ok
Image
btw whats an svc voice? any different from the regular rvc?
Yeah I still use so-vits-SVC
its that bad? 😭
svc is dead
oooooh
so it isnt replaying the voice changer its just putting out my voice
I go to the 3 dots and this is what appears 😐
and if i restart the space via the settings button, it still only shows this
It's loading
That's a good thing
It will stay like that for a few minutes
Also charge your phone
😭😭😭
Hmm, i see... i shouldnt be impatient 🤦♂️
Yeah I will
-rvc
RVC Guides (How to Make AI Cover)
Suggestions for @safe pike
Translation by country
thanks
trying to do
docker login registry.hf.space
docker run -it -p 7860:7860 --platform=linux/amd64 \
registry.hf.space/foo-rvc-v2-huggingface-version:latest python app.py
it has me login, that succeeds
then gives
Unable to find image 'registry.hf.space/foo-rvc-v2-huggingface-version:latest' locally
docker: Error response from daemon: unauthorized: authentication required.
See 'docker run --help'.
docker run hello-world works..
When I making a cover with a model maded by myselft I got the message "index -1 is out of bounds for axis 0 with size 0"
Someone know why?
trying to make a thingy with these instructions. i already have the model but how do i use the model???
https://cdn.discordapp.com/attachments/830489102620688407/1177849709717622784/Screen_Shot_2023-11-25_at_12.51.23_AM.png?ex=65740106&is=65618c06&hm=ef3f2a4cb43afe93847fc5457768f8fa4fd96e880747f741db7032daad0b5bee&
RVC Guides (How to Make AI Cover)
Translation by country
did you ever fix this?
doing the hugging face guide and wondering how do i use the model from weights.gg???
because right now it says 403 forbidden
403 means you're not logged in
where are you getting 403
hugging face
Ayo? @obsidian ravine level 1 !!! 
how to do tts?
console
i am using a collab but not workin
show me the logs
what colab are you using
i have reply on computer but it kills my laptop
i am trying to use replays trained model for tts
what model are you trying to use
i meant the model you're using
are there any tts that i can use my own model for ?
hang on @obsidian ravine
k
devi
si
worked
anyone know if there is a way to make tts with a model. i am looking at guides but i dont think i can upload a model
so wdym tts
text to speech
do you want to make tts models or tts audio to use with a model
tts audio to use with a model
i downloaded reply and took the model from there
idk if itll work but thats what i am thinking of doing. and then use that model for tts on collab coz reply kills my laptop
ty for your help!!
windows
try searching in C:\Users\username\AppData\Roaming\Replay\com.replay.Replay\models
replace username with your username
yea i found the model
but the part i am having trouble with is using the model for tts.
what colab are you using
i cant find a text to speech generator
ohhh
i started collab using this
cozi cant run local
no nvdia
i can train data and use the trained one model but cant the model i got . strange
try this https://ttstool.com/
do you know how i can insert the model from
try searching in C:\Users\username\AppData\Roaming\Replay\com.replay.Replay\models?
Ayo? @vague sapphire level 6 !!! 
show me the directory
Help pls.
Ayo? @long terrace level 2 !!! 
What dose the UVR-BVE-4B_SN-44100-1
do?
No idea
how can i use rvc in discord or can i?
-w-okada
-voicechanger
what cmd is this
-cmds
this?
Ayo? @small tangle level 1 !!! 
This interaction has expired, use the command
/guides realtimeif you wish to see it again.
😭
Fck
This interaction has expired, use the command
/guides realtimeif you wish to see it again.
This interaction has expired, use the command
/guides realtimeif you wish to see it again.
An epoch is basically how many times your dataset has been trained fully.
E.g. 200 epochs
Means the dataset has been trained fully 200 times.
what are the best gain in/out settings
and tune, index settings
That's voice changer but my gain is not changed and tune and index depends on you
Your voice
M->F 12
F->M -12
F->F/M->M 0
But it does still depend on your voice that's just the default that works on most people
depends on your mic lol
index, in case you wanna use it (because it consumes way too many resources), keep at 0,3 to 0,5
How much time is the Google colab offer a free runtime ?
So I used kits ai to try my voice model but for some songs the model sound was very low
How do I search Models
Ayo? @final tundra level 1 !!! 
...just don't use it
#🔍│find-models , type /search
Idk it's so simple but the limit is kinda bad
rvc is basically the same in terms of GUI lol
that's the thing
OK then
Mj
and duplicate that space
OK thanks for the tip
training rn
1 epoch is taking 22s
300 epochs with 15 min dataset seems good watchu think??
bro go for 500
??
alright listen up
i should really do that
so
kk
on a training session, your goal should be to hit certain milestones
500 epochs is the first milestone
check the graphs to see if it's overtraining.
is it good? hooray, continue
750
good? continue
1000
good? ...continue
if is not i just stop the cell right??
1250?
continue
got it
and then break the space time continuum
yeah basically
if ur training on colab try training for 1hr30 max
per session
to make sure u wont lose everything
it's simple
g e d??
then i just run resume cells in another colab or?
yeah resume cells in ur next training session
got it
yo guys what is best app to change voice for bad laptop
using rvc or not?
yes
idk but i know that u need a good pc 
ofc
check here #1159513888199540817
but wich one should i click
do you have an nvidia gpu
one that mentions training
i have gtx 1050 ti
not enough
does 3070 enough
oh
i got that card but im using laptop 
seems like ull need to use colab
sad
is colab bad?
no not really
i uploaded a model to hugginface but when i try to upload the model in rvc v2 it says error
-colab
- AICoverGen-NoWebUI, by Ardha Google Colab
- RVC Disconnected, Kit Lemonfoot Google Colab
- easyGUI, by rejects Google Colab
- RVC-HF, by r3gf Huggingface Spaces
- AICoverGen, by r3gf Huggingface Spaces
- RVC v2 Huggingface version, by Clebersla Huggingface Spaces
- Advanced RVC Inference, by neuclya Huggingface Spaces
You can find more info on the #1159513888199540817 channel. If you can't find your answer, feel free to ask for help in #✨│ai-help. Credits to Faze Masta and Antasma for compiling these links.
@slate gust 2nd and 3rd
ok
theres a guide for rvc disconnected
tnx
check here #1159513888199540817
where r u trying to upload it??
why when i use voice change it work but not work in call
so a w-okada situation
give me ur model link
lemme see if u copyed the right one
3070 is good already but you need to refer this for correct settings
https://rentry.co/W-Okada-FAQ
Ayo? @proud elbow level 9 !!! 
this is the pth download link
u need to put both pth and index in a folder and zip it
then upload that folder on hf
oh
any program that i can put only the zip file without uploading it to hugginface?
wdym
if u mean where to upload ur zip (alternatives)
there are many
for https://docs.google.com/document/d/1XuxQYiqEhYrdYeCZRRLrmV_ciMKo0bV-jTCGHu_-5Cc/edit
wdym by the dataset zip file (rvc disconnected)
drive, media, mega etc..
just zip ur dataset
and then upload it in the "rvc disconnected" folder that pops up in "my drive"
what does it mean by dataset like .wav's?
your wavs putted all in a folder
zip it then
alright
then do this
alright thanks
Ayo? @brittle wing level 1 !!! 
yw
-
/search command in #🔍│find-models
-
weights.gg (website)
-colab
- AICoverGen-NoWebUI, by Ardha Google Colab
- RVC Disconnected, Kit Lemonfoot Google Colab
- easyGUI, by rejects Google Colab
- RVC-HF, by r3gf Huggingface Spaces
- AICoverGen, by r3gf Huggingface Spaces
- RVC v2 Huggingface version, by Clebersla Huggingface Spaces
- Advanced RVC Inference, by neuclya Huggingface Spaces
You can find more info on the #1159513888199540817 channel. If you can't find your answer, feel free to ask for help in #✨│ai-help. Credits to Faze Masta and Antasma for compiling these links.
if u cant find there either it means that it doesnt exist
Thanks
how dose Ilaria UVR work?
What is the best voice changing app?
Ayo? @grizzled bane level 1 !!! 
#1159513888199540817 check here (owada)
just put ur audio
But what modles dose it use?
RVC Guides (How to Make AI Cover)
Translation by country
Why I'm getting this?
I instal Applio
but i don't under .. any think
any simple app can i use?
just click on the refresh button multiple times
ure on mobile??
no
cant really understand what this means
those in the guide are the one i know
Can someone tell me how analyse this graph
lmfao
what do i do here
someone please answer
guys
@copper pebble srry if this is annoying but can you help?44
lol
i made models when making models was easier
Ayo? @merry spindle level 5 !!! 
i dont usually use the easy colab
imo the easy colab is better than disconnected
its better
btw i think u need to paste your dataset path
where do i put the dataset.
audios is so i can test the voice
should i just use rvc disconnected since im KIND of more used to it?
whatever i will
u can use it
it literally creates a folder called dataset that u can put in the audios for training....
that doesnt exist
iN THE COLAB
im busyyyyy
you probably used different, older RVC disconnected, try the newer v2 one
https://colab.research.google.com/drive/1XIPCP9ken63S7M6b5ui1b36Cs17sP-NS#scrollTo=GpjFLdlRFlcA
no it's not rvc disconnected
that was easygui
im training right now everything is fine
it literally has a GUI which is impossible to use in colab free tier
but so far looks quite decent
actually quite possible, just has to be hidden
"hidden"?
testing it
What the fuck is going on anymore
Are we going full circle ?
From python implementation -> colab of the same thing -> eventually gradio (if that wasn't the case already) -> running on hugging face -> back to getting it to run from a notebook ?

.fm
The Girl Is Mine (Solo Demo)
Michael Jackson • Thriller - In The Studio

OwO i wanna mod it
Just use the normal uvr colab?
Whats the use of the gui
the one in hugging face?
No i meant why ilaria uvr on colab
Ok
lmfao
a fork of a fork of a fork??? 
Applio moment
by want you mean mod it 
Ayo? @hot lagoon level 9 !!! 
uhh is this good or bad progress? i'm using flac dataset like in the guide
Can someone explain how to analyse this graph?
it keeps 0%
Its fine
More to g/total graph
Move*
well...if you say so
@devout cloak
Ayo? @devout cloak level 2 !!! 
I don't get it
It works for me
I just downloaded it
If youre using an online GUI, just copy the link
If you're using a local GUI, download the zip file I just sent
@devout cloak
i am on iphone bruh
Im on samsung rn.
were you able to replicate the issue or does it just work?
I switched to mangio to no avail
Yes
Haven't tried yet
i don’t have one
...how
i think it’s a scam tho
NO FUCKING WAY 😭😭
Its not, it just sounds like iphone sucks for this
Wiat what
oh
Oh my fucking god
is rvc even a mod app?
Im blind
I'll just keep trying things, although im not sure what to try
let me search it up
These arent minecraft mods @devout cloak
they work tho
These are voice models for AI voice changers
Theyre not for minecraft
Bro...
How did you even get here
Who sent you here
i found it on yt
apparently you can get a villiger to talk
so i did
i want a villiger to talkkkkk
How old are you
…
Also link the youtube video
please watch the full video before doing anything
what video
.....
🤑
@devout cloak explain
explain what
i am appalled
Like what's the problem
i am flabbergasted
^^^^
Pardon me but what the fuck
No no it's not
may you help?
I'll explain it the best i can
It's not a mod, again, please finish watching the video you sent before doing anything
rvc is a voice modifying programm for computers (badly simplified)
rvc is not a minecraft mod / addon
voice models are supposed to be used with rvc
voice models are NOT supposed to be used with anything else (especially minecraft)
^^^^ this video ^^^^
It's not related to Minecraft at all
i watched where the villiger talks
It's a seperate thing called AI voice
and how to get it to talk
therefore you didn't watch the full video
how do i get that
yea but i didn’t need to
Basically you use ai to sound like a villager or anything else via using certain models
it's not a mod, it's a program for PC, not mobile, that allows you to convert voices to other voices
all i needed to hear is the villigers
I- ...what?
Actually wait
talking
Are you on pc?
no
Just tell me I'm curious
he's on iphone
just so that I can adjust how to explain stuff, may I know how old are you?
Please try to understand what clip is trying to say
Yo
Sup
@devout cloak
-
The videos the guy makes is staged, he uses AI to pretend the villagers are talking but its edited
-
The video he sent teaches you how to use AI to sound like a villager, not how to make the villagers talk
-
This is using the voice changer software he linked in the video description. At NO POINT do you modify minecraft game files or put any villager.zip files into minecraft
^^^
additionally, you can't do something like this on a phone (or well, you can but it's not intuitive, especially given your apparent skills on tech).
Just so you know
this is why i am not a helper anymore
still surprised about how much patience you guys have
i would've already snapped and started being snarky sarcastic trolling piece of shit
lmaoooo
it's hard but you gotta have that patience because one day you could have the same questions
in a different thing
how do i make minecraft rocks talk
^^^^
if you would have watched the first 30s of the video you wouldnt need our help rn
I blame the youtuber who fooled them
This kid is 7, theres no point snapping
lol
he didnt even watch the fucking video and then complains
Also this is very entertaining ngl
30 SECONDS
Wouldn't be surprised if he was underage 
i'm afraid of children
plus, they somehow manage to frustrate me by simply talking
Hey now you were a dumb kid too once upon a time
No not really lmao
but im sm0rt now
I used to watch minecraft mod showxcases until my brain went numb
I was at a 6th grade reading level in 2nd grade 

in my days minecraft modding was never invented sadly
well, kinda
And now look at you
nowadays i can't play minecraft without at least 40-50 mods
Yeah same
Those mofos in my basement can't stop crying
Like come on just lemme do business
....there's children in your basement...???
btw guys who wanna come watch me suffer in minecraft's modpack gregtech new horizons?
(related to rvc 100%)
real
There aren't any in yours..... ???


