#🧬│ai-chat
1 messages · Page 353 of 1
who has a python script that can split audio into ten seconds? i had one but lost it
'
use free batch music splitter
link bro? thanks
applio has a method in preprocess that slices into 3 sec chunk and 0.5 s overlap (can be adjusted)
maybe isn't for RVC, RVC will split the audio anyways
its for appilo
anybody know how to use AI to do parody songs? Like have a voice model sing a song with different lyrics but with the same tempo/style/notes as the original song? Like what Weird Al does but using AI?
its very important to say this option is only meant to be used in in datasets where silence has been truncated
otherwise it would lead to horrible slicing
personally I never use built-in slicer. the default -40dB gets rid of breathe details
if the truncated silence is too short, applio inference may leave random noise on silences
yes thats why you have to leave 0.25 seconds of silence in every chunk
this and using 3 sec and 0.3 overlap is safe
u can play with different overlap/chunk values but that requires leaving a different amount of silence in the dataset
^ dis one has been proven that works fine for 3 sec and 0.3 overlap
is there any python package that can draw llm model architecture graphs. except torchview and graphviz?
If you wanna promote your portfolio, go to #1159290752195633273.
@covert lake @tepid basin
@paper blaze удали сообщение пж
damn, that's crazy
is there really no one who can delete that message currently
oh there we go
huh
is the voice changer safe?
Yes as long as you download it from our official links
-realtime
Interaction has expired, use the command again for a new interaction.
-rvc
- How to use RVC Mainline Colab by Cauthess
- AICoverGen Colab Guide by Eddy (Spanish Helper)
- Create a model with RVC disconnected (colab) by Angetyde
is this copyrighted
depends if you think suno ai is copyrighted lol
help kind people, where can I download the voice module for Voice Changer
what are you listening to
the opposite of nightcore
-rt
Interaction has expired, use the command again for a new interaction.
any working colab notebook to train rvc models ?
-colab
Suggestions for @plain sandal
- Applio, by IA Hispano Google Colab
- RVC Disconnected, by Kit Lemonfoot Google Colab
- RVC Mainline, by Hina Google Colab
- AICoverGen-WebUI, by Hina Google Colab
- AICoverGen-NoWebUI [English], by Ardha, fixed by Eddy, Hina and Gdr Google Colab
- AICoverGen-NoWebUI [Spanish], by Eddy, Hina and Gdr Google Colab
- UVR5 NO UI, by Eddy Google Colab
- UVR5 UI, by Eddy Google Colab
- Modified W-Okada's Voice Changer, Google Colab
- FaceFusion UI, by Nick088 Google Colab
- FaceFusion NO UI, by Nick088 Google Colab
- EasyGUI, by Rejekts Google Colab
While the Colab free plan provides up to 12 hours of daily usage, the GPU is typically available for only about 4 hours each day on average.
AI is great I can finally hear Nas rapping Gettin Jiggy Wit It
oh dear
wanted something that has resume training support, that was a customised notebook but it is showing an error now
https://colab.research.google.com/drive/1PorBmm4bUjY1-dk4f7XKrO7Mx51wbqYU#scrollTo=w4wXvoez9Rce
here it is if someone can fix it
It's clearly mentioned on the notebook that it is banned for free users
Use RVCdisconnected
outdated notebook not recommended
pls br server voice model convit
what happend to sam the demonic announcer its gone everywhere
how can i list all subdirs under this directory? https://myanimelist.net/profile/
i want to create a dataset for machine learning
i need profile names
1st one is better
-colab
- Applio, by IA Hispano Google Colab
- RVC Disconnected, by Kit Lemonfoot Google Colab
- RVC Mainline, by Hina Google Colab
- AICoverGen-WebUI, by Hina Google Colab
- AICoverGen-NoWebUI [English], by Ardha, fixed by Eddy, Hina and Gdr Google Colab
- AICoverGen-NoWebUI [Spanish], by Eddy, Hina and Gdr Google Colab
- UVR5 NO UI, by Eddy Google Colab
- UVR5 UI, by Eddy Google Colab
- Modified W-Okada's Voice Changer, Google Colab
- FaceFusion UI, by Nick088 Google Colab
- FaceFusion NO UI, by Nick088 Google Colab
- EasyGUI, by Rejekts Google Colab
While the Colab free plan provides up to 12 hours of daily usage, the GPU is typically available for only about 4 hours each day on average.
this some heat or nah https://www.youtube.com/watch?v=MEfldrmSoUg
How do people make those ai picture videos
Dunno if I asked before but I'm looking for a RVC tool or anything alike that can batch convert multiple audio files with a selected voice model
-colab
- Applio, by IA Hispano Google Colab
- RVC Disconnected, by Kit Lemonfoot Google Colab
- RVC Mainline, by Hina Google Colab
- AICoverGen-WebUI, by Hina Google Colab
- AICoverGen-NoWebUI [English], by Ardha, fixed by Eddy, Hina and Gdr Google Colab
- AICoverGen-NoWebUI [Spanish], by Eddy, Hina and Gdr Google Colab
- UVR5 NO UI, by Eddy Google Colab
- UVR5 UI, by Eddy Google Colab
- Modified W-Okada's Voice Changer, Google Colab
- FaceFusion UI, by Nick088 Google Colab
- FaceFusion NO UI, by Nick088 Google Colab
- EasyGUI, by Rejekts Google Colab
While the Colab free plan provides up to 12 hours of daily usage, the GPU is typically available for only about 4 hours each day on average.
Mainline and EasyGUI aren't working for me 🤔
gui banned for free users, i was using it 2 days ago fine
Idk
Hello everyone, is there any AI that can automatically choreograph dance videos based on music and pictures or 3D models, and the output is similar to mmd
Hello here too,
I think Hina_Mod_AICoverGen_colab.ipynb is down (gradio) ?
The public (and local, which I don't use) link is no longer displayed.
there should be something alternative, i don't have GPU 😦
If y'all wondering why any Python notebook is broken on Google Colab, the #📰│dev-updates has explained. If you don't wanna wait and you wish to use Google Colab anyways, there's a way you can downgrade Python to a specific version right before you run any pip or install command.
However, this method is currently in research. I will add more commands to this if there any issue. 
Ok, great. The pip command from Python 3.10 now works.
anyone know of any video models that can be run locally ?
using uv to install 3.10.12 on venv seems to work too
@stark scarab
So, trying to get Colab to use Python 3.10.12 as its main Python version is a bit complicated. A lot of packages are missing from Python, which is installed from start, including the pip command and especially Google ones. The only one package I couldn't pip install is google-colab, which I think this package is critically required for everytime you start or restart the session. After you do !sudo apt-get install python3-pip, you must upgrade its pip version to the latest before doing any pip package install, because it gave me pip 22.x instead of pip 24.x. Despite all of these, it somehow worked fine. 
To #1159290752195633273 your thing, go to #1159290752195633273.
doesnt seem to work with kaggle
Kaggle and Colab work different, as what I know.
yup but for the most part they are both using jupyter
I have the AI voice changer but it keeps making a glitch sound with the voice
Hi
I wonder why there is no "Zant voice" in every "ai voices" this is.. dissapointing.. T-T
@naive nymph паже разбань
Whats your GPU?
even if you have a 4090, it won't be faster than old school 3d pre-rendering ||(just a rough assumption)||
One message removed from a suspended account.
:wave: @terse bolt, How can I help?
Available Commands:
• @weights find <query> or /find <query> - Search for RVC Voice Models
• /create - Create an AI Cover
• /image - Generate an Image
To use Weights bot, go to #🤖│bots or #🔍│find-models.
I've been playing around with weights, and I cloned my own voice, I fed the thing recordings of my own voice from songs, without the effects, just acapllas, and several contain screaming and growling, but for some reason if I try to make the model sing a metal song with screaming it does those weird artifacts that HHHHH, why is this? What's the workaround?
RVC idea:
Safetensors RVC
?
heyy does anyone have a good real time modulator?
There's no workaround.
RVC's pretrain limitations are the reason.
Wake up, everyone! A new RegalHyperus drum model just released!
The Rising Fighting Spirit (Drum model no. 572)
Hey , I recorded a video and I need a voice changer for the audio of the video
Any recommendations , what can I use
as voice changer for already pre recorded audio?
that's what RVC does
soooo does anyone have an alternative for rvc disconnected?
Kaggle or Lighting Ai
are these colabs?
send link please, i'll try it🙏
https://docs.aihub.gg/rvc/cloud/applio-kaggle/ and https://docs.aihub.gg/rvc/cloud/applio-lighting-ai/
Last update: Jan 13, 2025
Last update: Nov 21, 2024
RVC as Retrieval-Based Voice Conversion, not the Realtime Voice Changer
@tacit nimbus you're asking a silly question.. RVC works for all languages
elo
Guys is the deiteris fork still safe?
Just tried to download it on a new pc but it had a different download host compared to before. Also Virustotal flags it with 6 issues while the old file only has 4
where are u downloading it from? share the link in #🔍│help-w-okada along with ur pc gpu
Кто дс, кидайте в др
Unfortunately, my birthday has always been the lousiest day of the year. It all began on the day of my actual birth... Both of my parents failed to show up. By the age of 5, I was forced to throw my own surprise party... at Gunther Goat Cheese's : The goat-cheesiest place in all of Druselstein. Many of my closest friends were there: Count Wolfgang, Betty the She-Boar, Ratputin, and the licekins, Olga and Chicago Joe. But one should never walk the paths of Druselstein. With uncovered Doonkelberry cake, lest the Doonkelberry bats swarm.
I have a 4070
Damn really ?
Are you planning to Generate high Quality Videos?
Thats too bad i figured if we can generate images then wed be able to generate video one frame at a time
Id be happy if i could just make low quality
Im not too picky
Max Video lenght is 5 Seconds most of the time
Just so you know
making some animated gifs is more viable ig
Damn really ? There isnt any way to do it bit by bit in tiles ?
What can i use to make those 5 second vids ? Comfy ui ?
I never used comfy UI
Id be interested in that too actually
Help
Is there an alternative to character ai with the "guidelines" that character bots have? I would continue using character but there have been a few things occurring that make me uncomfortable with using the app.
hello, someone could help me on how to use a Miku Ai Voice Model on real time voice?
hello
Is there an alternative to gen-3s video uncrop/zoom out
Or audiosep (separate what you type)
yup for stuff that uses uv, you only need that and everything should work
Hi guys, I'm a girl who wants to sound like a guy. Any recommended model?
I call cap on the first part but to find male voices you just just scroll through #1175430844685484042 or look around in weights.gg
yeah but I was wondering if anyone else had any model they specifically recommended. Been trying with bad results
the best bet is to try several voice models since only you can judge if the quality satisfies you or not
Кто пойдет в дс? Кидайте приглос в др
You might like something
https://www.weights.gg/user/clncvzj57000zmf08xpzp4hm2
ryou_yamada on Weights | Check out models and comments by this user.
hi there, i have any question.
is there new colab notebook for run rvc ? thanks
How is this going btw? is it stable enough to get added in #📰│dev-updates
CalledProcessError: Command '['pip', 'install', '--upgrade', 'setuptools']' returned non-zero exit status 1. gives me this error the colab rvc disk after phyton your : @solar torrent
I'll mark it for you
in what sense?
You don't know how to !python3 -m pip install --upgrade pip? What?
Tried that earlier on voice changer, it worked on some packages and some didn't install at all
Namely colorama and pyngrok from what I saw
It says it's installed but can't be found my the code
No module found or something
Hey, I was using an RVC on Google Colab
BahaaAI, very easy to use interface
but since yesterday it stopped working
Read #📰│dev-updates for more information.
Google has updated the main version of Python to all Colab notebook.
Na more like colab updated and some stuff broke, including rvc stuff
ok, thanks a lot
I was kinda lost
good to know there's a lot of people who are aware of the issue here
Yup if you're trying to train or do inference (change voice from song) mainline is currently functional
Doesn't have auto remove background tho
Yup
they give you 1 hour free per day
Oh so you should be fine then
but I always have multiple accounts
Damn that's painful
🤷♂️
I make YouTube videos, no way I'm doing 1 hour a day
it's more than enough
but their plans are weird
1 hour free, 4 hour for paid version
like, no, I'm not paying for just 4 hours?
When stuff are back up you can use UVR colab to remove background for free
I will just make 4 Adobe accounts 😂
Or use msvep online for free
they don't even check IP
Pretty sure this doesn't have a limit, it only has a queue
code? Thanks 🙂
12-24 hours
no
I mean
until the updates that broke stuff get fixed
or does it requires every Colab project to fix its issues individually
Ye it requires manual fix for every single one
Unless google somehow gives an option to downgrade python version
after doing that it works fine?
it's just:
!sudo apt-get install python3.10
!sudo update-alternatives --install "/usr/bin/python3" python3 "/usr/bin/python3.10" 1
!sudo update-alternatives --install "/usr/bin/python3" python3 "/usr/bin/python3.11" 0
!python3 -m pip install --upgrade pip
right?
nope just an update who broke previous code, check #📰│dev-updates , there's alternatives
Good morning guys
Yes. If one !sudo update-alternatives --install "/usr/bin/python3" python3 "/usr/bin/python3.1x" ends with 1, this Python version will have more priority and the Colab will pick this version to use. And if one !sudo update-alternatives --install "/usr/bin/python3" python3 "/usr/bin/python3.1x" ends with 0, this will make this Python version to have less priority and not be used with Colab.
what colab did u test it with? was it stable/working?
Both original demucs and the Deiteris's fork W-Okada Colab notebooks. But I haven't yet tried this with fork W-Okada Colab.
Before I test anything with Colab, I will set Colab notebook to use CPU first and before T4, just so Colab won't be wasting too much time usage while running. 
Fork doesn't even need python to run since it runs a prebuilt version, so it shouldn't be affected
How install pytorch??
PS C:\Users\-_-\Documents\проэкт\snake-ai-pytorch-main> pip3 install torch torchvision torchaudio ERROR: Could not find a version that satisfies the requirement torch (from versions: none) ERROR: No matching distribution found for torch PS C:\Users\-_-\Documents\проэкт\snake-ai-pytorch-main>
Try use pip instead of pip3.
ok ty
C:\Users-_->pip install torch
ERROR: Could not find a version that satisfies the requirement torch (from versions: none)
ERROR: No matching distribution found for torch
C:\Users-_->
could u test and let me know please
ok, I managed to get it to work with Applio
thanks for everything everyone
that was a lifesaver
what did you do to manage to make it work?
did you use the script namari said?
Worked fine for me, even with Python with already installed packages. Unless the website that stores Python package is blocked in your area.
I just used Applio using a tutorial 🤷♂️
tho I used inference
I already had the model (just had to convert it to zip)
wait lemme check it out if it has been updated or used another way to use python
use a specific version and or extras for cuda
pip install torch==2.3.1 torchvision==0.18.1 torchaudio==2.3.1 --upgrade --index-url https://download.pytorch.org/whl/cu121 for example
im getting this whenever i press on any model and i do have pytorch installed
Traceback (most recent call last):
File "voice_changer\RVC\embedder\EmbedderManager.py", line 44, in loadEmbedder
File "voice_changer\RVC\embedder\OnnxContentvec.py", line 17, in loadModel
File "onnxruntime\capi\onnxruntime_inference_collection.py", line 347, in __init__
File "onnxruntime\capi\onnxruntime_inference_collection.py", line 384, in _create_inference_session
onnxruntime.capi.onnxruntime_pybind11_state.InvalidProtobuf: [ONNXRuntimeError] : 7 : INVALID_PROTOBUF : Load model from pretrain/content_vec_500.onnx failed:Protobuf parsing failed.
During handling of the above exception, another exception occurred:
Traceback (most recent call last):
File "voice_changer\RVC\pipeline\PipelineGenerator.py", line 30, in createPipeline
File "voice_changer\RVC\embedder\EmbedderManager.py", line 25, in getEmbedder
File "voice_changer\RVC\embedder\EmbedderManager.py", line 48, in loadEmbedder
File "voice_changer\RVC\embedder\FairseqHubert.py", line 11, in loadModel
File "fairseq\checkpoint_utils.py", line 425, in load_model_ensemble_and_task
state = load_checkpoint_to_cpu(filename, arg_overrides)
File "fairseq\checkpoint_utils.py", line 315, in load_checkpoint_to_cpu
state = torch.load(f, map_location=torch.device("cpu"))
File "torch\serialization.py", line 797, in load
with _open_zipfile_reader(opened_file) as opened_zipfile:
File "torch\serialization.py", line 283, in __init__
super().__init__(torch._C.PyTorchFileReader(name_or_buffer))
RuntimeError: PytorchStreamReader failed reading zip archive: failed finding central directory```
thxx, btw I would just suggest to change some text since it looks like the original one, like in the title put that it's the deiteris fork
btw, good job with the fork, people reported that they also had better quality
Seems to work on RVC Disconnected, but you need to install tensorboard manually for some reason
Hi all, I am not sure if here is the right place.
But Does AI TOPS Impact AI Inference and Training Speeds?
Comparing GPUs AI TOPS from Nvidia website
Feature AI TOPS
RTX 5090 3352
RTX 5080 1801
RTX 5070 Ti 1406
RTX 5070 988
RTX 4090 1321
RTX 3090 285
from my understanding, 4090 shouldn't be 4 times of 3090 in ai training and interface
i think it includes only tensor core performance
so if some ai uses only tensor cores then it'll be way faster than 4090
rvc, if tf32 precision is enabled, also uses tensor cores to some extent
first, what AI application? it may use different precision, hence the performance varies
also 5080 fp4 > 4090 (fp8/fp16?) is not a fair comparison
personally I am use gpu to local llm /TTS /STT for game mods e.g. chim and desktop Masuku with rtx 2060 6 gb
I am considering to upgrade my gpu
that sounds like a marketing gimmick, like saying 5070 (with DLSS4) is faster than 4090 (native)
then I saw some small tech youtuber just pull this info to compare which I found irrelevant
4080->5080 is like 10% diffrence in gaming (without dlss)
yeah I think its a market gimmick
but I'm curious about how big the improvement is for these new GPUs on generation time local llm.
well yea the blackwell architecture introduces accelerated fp4 computation, where last gen cards don't enjoy the benefit
would a rtx 5070 much faster than rtx 3090 while they are similar price
@patent lake
or try uv
the former has half vram, whereas 24 gb vram could allow running flux non-quantized/reduced
Currently at work, will look into later
pls @ me when u fix it
CalledProcessError Traceback (most recent call last)
<ipython-input-1-54e893f947a1> in <cell line: 0>()
19 for package in pip_packages:
20 print(f"Installing {package}...")
---> 21 subprocess.check_call(['pip', 'install', '--upgrade', package])
22
23 # matplotlib version force to fix tensorboard error
/usr/lib/python3.11/subprocess.py in check_call(*popenargs, **kwargs)
411 if cmd is None:
412 cmd = popenargs[0]
--> 413 raise CalledProcessError(retcode, cmd)
414 return 0
415
CalledProcessError: Command '['pip', 'install', '--upgrade', 'faiss-gpu']' returned non-zero exit status 1.
anyone know how to fix
in what sense install it manually ? ( mini tutorial for everyone to get it working )
use help channels, use #✨│ai-help and link the tutorial/colab u are using
!pip install tensorboard
so I only have to put this in and that's it?
After everything showed by Namari up there yeah
ok i will try it
I tried it works the key la the most important one let's say always gives error
Command '['pip', 'install', '--upgrade', 'setuptools']' returned non-zero exit status 1. to me appear in front of me always this error how do I do? (even with the things you gave?)
Still this lol, the only reason I actually want to keep a filter is because I'm not 18+ so I don't want the AI bots trying to make a move on me 😭🙏
@twilit peak
Try lowercase u
no it is right but I don't know why it gives me error
So weird
I never had any of those errors
Maybe try disconnecting the runtime and start anew
ok
maybe try sending a mini script to put on top, could be useful as a temp solution
So when I used the new uvr update on the site when I tried to use the de reverb it took away doubles on audio tracks! Even made some acapellas and instrumental. And even made some tracks sounds disappear and only have the baseline with the vocals.
Crazy .
I wonder if Baldi sings Pony from Uncle Samsonite.
Do you have a modifier file here?
wdym?
pdf file
do i have sexz here
oh dear
I use play.ht
You upload Acapella. Voice transcription
i think i can immitate biggie smalls voice good!! this was a test
nope
what biggie smalls waffle all about 💀😭
Dunno just seeing if my white voice can sound like him trying to impersonate him before the rvc sounds pretty good tbh!! Might try to do a jr mafia song he wrote for them as a test
For Inference (use models) Mac, You can:
- Locally (runs on your pc so the speed depends on that, you will have to set it up with the guides, probably won't be able to train, make models):
- Cloud (remote good pc, easier and faster than ur PC but it's limited):
- Weights.gg: Partnered with AI Hub, lets u do them easily but u may be in a queue
- Ilaria RVC Zero: fastest and simplest that you can get for free
- Applio Colab: max 4 hours, not granted, of GPU
Just use your phone and send to pc that’s what I do
I record on my phone the. Send file to pc
but if i only have a Mac
i dont have a pc
do any of you guys know any good guides on how to set up ai voice changers?
Just buy a prebuilt nvida that’s what I got had since 2020 2060 super
it seems more of rvc inference
and it doesn't seem a cloud service
Unfortunately it is not open source. You know w-okada depends on Pytorch still even ONNX mode. We developed Pytorch free implementation.

It realizes low VRAM usage and works on Radeon.
if it runs locally, ideally it should be able to run completely offline
no internet connection nor sort of DRM shit needed
plus, it uses rvc, not your own proprietary architecture. if it is latter, you could patent it and sell the software as well (for instance, vocaloid & synthV).
@naive sonnet don't post the same stuff in multiple channels
sorry
Have you ever find one in #1175430844685484042?
Hello guys, I need help about voice changer client please
!howtoask
How To Troubleshoot 
- Don't simply mention your issue, like "
my rvc is not working". - Describe the step you are on, what you're trying to do, the RVC you're using, a screenshot, etc.
- The more context, the better.
- Don't be desperate. You can ping a Helper, but if they ignore, they aren't available/don't know the answer.
- It's okay if you're frustrated, but don't take it into this server.
- Don't DM without prior consent.
- Don't ask for every little instruction. Put your own effort & test things by yourself.
- Don't ask to ask.
- Check if your answer is a Google search away/on our guides website.
also go to #🔍│help-w-okada and read the pinned guide there
The chat isn't where you ask for a download link of W-Okada. W-Okada isn't that closed secret. 
and how do i do it on Mac
how i do import that mdoel into uvr
anyone knows the name of the ai voice in this vid? https://www.youtube.com/shorts/psIgF-XtNBc
not the mickey mouse one
the other one
I feel ugly.. I made my model, but it's messy.. I tried to use a site to use RVC and here I am.. it's a mistake, I tried everything but I feel that destiny is against me..
why do you type like this
Because.. it is used of me.. I love writing such as this I dunno why x3
typing quirkcs.. grrr
we accept ONLY english in this server unless in language channels
Wake up, everyone! A new RegalHyperus drum model just released!
Thunderings of the Merciless (Drum model no. 573)
any good female voice models?
Guide style is in the same as vtarcelia. Thanks vtarcelia for corrections. Most technical information comes from deiteris.
Last update January 17: NEW UPDATE VERSION b2332 (from December) , adjusted known settings
Translations added for:
German: https://rentry.co/ForkVoiceChangerGuide_de
Turkish:...
You can search rvc ai voice models at:
- #1175430844685484042
- In #🔍│find-models , Do /find with @hidden grotto
- https://weights.gg/ (login required)
- https://huggingface.co/models (but watch out cus in hugging face there arent only rvc ai voice models)
- https://voice-models.com/
- https://thevoicemodels.com/ (for Turkish Models, login required with discord and level 2 on their server)
if there isnt one, you can:
- #1159289738314919936
- #1191429836321849435
- make it yourself with our docs guides https://docs.ai-hub.wtf/essentials/how-to-make-voice-models/
:wave: @covert lake, How can I help?
Available Commands:
• @weights find <query> or /find <query> - Search for RVC Voice Models
• /create - Create an AI Cover
• /image - Generate an Image
aaah...
For the real time voice changer what’s the best settings for NVIDIA GeForce GTX 1070 (8gb)?
Imagine asking for the same question in every channel, but still no answer or response. 
Hlw
kakera
What's the best model on ultimate vocal remover for clean instrumentals?
waht
im on the voice changer client and it dont make me sounnd like it
it dont work
the voice changer
Why the hell are you asking me?
Don't ping some random user here. Not everyone knows how to install W-Okada.
@oblique creek
lol, not being able to install w-okada when there's literally a guide for it

@bold lily
lol, delivering drugs?
how rich sugar mommy
pay me first before i start paying you as proof
Selamün aleyküm arkadaşlar ben oyun konsol
can i ask a basic simple doubt here?
or can someone point out the right place?
how to verfiey this server
dont ask to ask
just ask a question
!howtoask
How To Troubleshoot 
- Don't simply mention your issue, like "
my rvc is not working". - Describe the step you are on, what you're trying to do, the RVC you're using, a screenshot, etc.
- The more context, the better.
- Don't be desperate. You can ping a Helper, but if they ignore, they aren't available/don't know the answer.
- It's okay if you're frustrated, but don't take it into this server.
- Don't DM without prior consent.
- Don't ask for every little instruction. Put your own effort & test things by yourself.
- Don't ask to ask.
- Check if your answer is a Google search away/on our guides website.
I wanted suggestions for a from-scratch implementation of Knowledge Distiller
I was following a knowledge distillation course from YouTube
I could hardly find any course which showed a practical hands-on approach for it
So, I only implemented the one project that I could find for distilling: bert-base-uncased-finetuned-clinc → distilbert-base-uncased
This is the link for it
https://www.youtube.com/watch?v=mY12NlLxBi4
Here's my very slightly modified version of the code notebook for it
https://colab.research.google.com/drive/14_8I8zgeuSrHAwBsinDy2SMoTnahtgnY?usp=sharing
What else should I add to this code for a better code? Any extra features/implementations that I can add to make it better, or to enrich this code?
Please suggest, thanks
@covert lake hey there nick how are you? idk if u remember me or not so anyways can you help me with something
I'm good what's the matter
well do you remember back then struglling with the Realtime voice changer?
Not much it's months ago
There are different Text To Speech (TTS) AIs:
GPT So Vits: RVC isn't as good as GPT So Vits for tts, but gpt so vits (few shot tts, which means needs just a lil training for models) can't use rvc models (and viceversa), and its only limited to: english, chinese & japanese, if you wanna check gpt so vits instead, read https://docs.ai-hub.wtf/tts/gpt-sovits/
Freemium 11labs: An easy way to do TTS is https://elevenlabs.io/, you can't use RVC model on this but its a mostly premium easy way for good quality TTS
FishSpeech: FishSpeech is a 0 shot (no explicit training needed) TTS, if you got a good pc you can use it locally else use their site
With RVC Models:
RVC is natively for Speech To Speech, but forks such as ilaria rvc mainline & applio have built in tts (using Microsoft Edge TTS to make a generated tts audio, which i suggest you to choose a tts model that is the same gender and language of the rvc model you wanna use, and then convert it with rvc)
If you wanna do tts locally with RVC Voice Models (if you got a good pc):
If you don't got a good pc you can do tts with RVC Voice Models on cloud:
-
Ilaria RVC Zero (Running on A100 GPU, free fasted rvc on cloud) and the guide
-
Use Applio UI Colab (with google colab T4 free daily limit gpu)
-
if you don't wanna use edge tts, you could try another tts ai from our tts index and use the output as an input in rvc
You can do this for calls but it's not really realtime
ik i want it as audio files
Yeah u can also follow that I said above
Read it
Then try Ilaria RVC Zero or Applio UI colab
However RVC is not made for tts, it won't be emotional
aw dang it
i want a screaming audio for metallix
Prob can't do that well with TTS
You can try 11labs or GPT-SoVITS or F5TTS or Fish Speech, but TTS ain't that good with screaming
well why wasting my time if its not good
You can try, but AI isn't good with screams
Does anyone know why I put a .wav in applio, it doesn't detect it?
anyone know any free ai video software
in google colab
GPU Nvidia T4=is good for use IA already trained
TPU Google v2-8=is good for train IA with PyTorch or TensorFlow
CPU=little IA (example:differentiate if the image given is a cat or a dog)
HaiperAI,FlareAI,Luma AI,Runway and hailuo
Any JavaScript developers want to collaborate on a project?
anyone who can help me with pre recorded audio to ai voice?
on mac M1
I dont understand how to make it work
when I try to install applio it wont let me on terminal
is there a way to just do it online ?
i mean... you got mac, your only choice to make a model is by using google colab or kaggle
kaggle is a bit harder to setup, but should be faster
i dont find anything about it
-colab
- Applio, by IA Hispano Google Colab
- RVC Disconnected, by Kit Lemonfoot Google Colab
- RVC Mainline, by Hina Google Colab
- Hina's Mod AICoverGen WebUI, by Hina Google Colab
- AICoverGen-NoWebUI [English], by Ardha, fixed by Eddy, Hina and Gdr Google Colab
- AICoverGen-NoWebUI [Spanish], by Eddy, Hina and Gdr Google Colab
- UVR5 NO UI, by Eddy Google Colab
- UVR5 UI, by Eddy Google Colab
- Hina's Modified Original W-Okada's Realtime Voice Changer, Google Colab
- FaceFusion UI, by Nick088 Google Colab
- FaceFusion NO UI, by Nick088 Google Colab
- EasyGUI, by Rejekts Google Colab
- 🆕 Music Source Separation Training (Inference), by Jarredou & Makidanye Google Colab
While the Colab free plan provides up to 12 hours of daily usage, the GPU is typically available for only about 4 hours each day on average.
-kaggle
- Applio Notebook, by Vidal Kaggle
- Applio Notebook, by Shirou Kaggle
- Music Source Separation, by Shirou Kaggle
- UVR5 NO UI, by Eddy Kaggle
- Original W-Okada's Voice Changer, Kaggle
- Modified W-Okada's Voice Changer, Kaggle
- 🆕 UVR5 UI, by Eddy, ArisDev & Nick088 Kaggle
- 🆕 RVC AI Cover Maker UI, by Shirou & ArisDev Kaggle
- 📖 How to use RVC Mainline on Kaggle by Cauthess
Note: Kaggle limits GPU usage to 30 hours per week.
dont see why not
which one of these links
not necessarily a song
just a voice recording of someone speaking
chill, he has been bonked already
wait ur the person who made koe.ai?
No, I'm not the person. No relation.
oh alright i was confused by the name of your product "koemake"
thank you for clarifying
No problem. Right. Similar name.
hi
@severe crown Koemake has a built-in virtual audio IF! Wonderful!
We don't allow this type of promoting here, it's not the first time someone comes here saying they made a better paid wokada, and it can't even be proven it's PyTorch free since it's not open source
I got a better "RVC" voice model voice changer, it's called Detris' fork W-Okada, and y'all still advertising your "paid" service for me to try huh. 
oh very closed place.

The whole app is closed source and all of the messages are just trying to promote it for some money, which is why I said to use #1159290752195633273 , it's the right channel for that
Deiteris >>>

Money leeches from a free software. 
Just make it freeware-open source, bros, and I'll give it a try.
that's funny

Hello guys
Hewwo
Wassup
I'm otay hbu?
I'm doin great
What's the best app for voice changing
W-Okada.
You still ain't over with this one? Lmao.
You admitted the service isn't free. If you want more people to use your service, just make it completely free for use. Don't limit the features to those who haven't paid for it just because it's not free. The only paid parts should be your merchandise products and way to donate. 
This is my suggestion. But if you still wanna go for this way, your service ain't gonna boom-boom either. 
Not my service. I’m just a user of it.
You are disagreeable about everything are so annoying.
wokada
#🔍│help-w-okada
tell what's ur pc gpu, what tut link u followed and what model u wanna add
help wokada is the right channel
Don't randomly ping people
tell me ur pc gpu in #🔍│help-w-okada , I'm guessing realtime voice changer
Don't suggest users to use a paid program when there's a free one
realtime voice changer for calls? tell me ur pc gpu in #🔍│help-w-okada
@nocturne finch Talk only english here, and we don't allow any type of NSFW.
Don't talk in any languages other than english here.
I suppose he cant speak english 
ima deal with this rq
Hi friends

Hello Everyone, hope y'all are having a good day?. Can anyone make me a good female voice message for free? Short messages. Please, Thanks!!
Enough to piece a voice scene together? DM, let's talk.
i am testing a dougdoug model and i need an audio test liek basic talking
Can i listen to the model?
i need a audio test first
then il be able to send it in dougs voice
I believe doug is male? Yes?
DM let's talk about the audio test and know how to work it out..
why exactly
eh
ok
Why what?
why do we need to dm about it
Chatting about it here is alright? I'm not particularly familiar with the chats here...
Oh okay. Just a few more periods. Nothing to it.
So? Being a free service, is that a problem? Lol.
now shush cuz the admin or mods is gonna ban if u dont stop o - o
that guy is so clearly paid to promote that stuff or even is the alt of the creator of that thing lmao
You guys ever implement deepseek locally? Any hurdles?
how do i train a model
4060ti 16gb vram yeah i wanna a custom to use in rvc or tts
Twin lol
Well RVC is STS natively even if it can be used for TTS
thank you for the info
yw
@ionic folio
what
i have always been here before you
bro u are not my future self
I think bro is ur lost twin 🔥
@chilly lake i have a question. is there a feature in applio's 3.2.8 bugfix update where you could select which vocoder to use to train a model?
(sorry for bothering you)
bc i really wanna like, try out the klm 4.4 mrf-hifigan pretrain
clone the main repo instead of the release 3.2.8, or try codename's fork
i train on colab ;(
should i remove the "--branch 3.2.8-bugfix --single-branch" from the "!git clone --depth 1 $uioawhd $new_name" code?
you need to clone the repository for that
just remove --branch 3.2.8-bugfix
how dose one make a ai cover
There are many ways to do AI voice cover.
okay if i want to do voice to voice genaration how would i
the fact u have that on hand says alot about you 😆
eleven labs from what i can understand dose it but only with there own trained voices
-docs
Suggestions for @cedar tapir
- How to use RVC Mainline Colab by Cauthess
- AICoverGen Colab Guide by Eddy (Spanish Helper)
- Create a model with RVC disconnected (colab) by Angetyde
-rvc
- How to use RVC Mainline Colab by Cauthess
- AICoverGen Colab Guide by Eddy (Spanish Helper)
- Create a model with RVC disconnected (colab) by Angetyde
oh um same thing
im downloading rvc rn
Ai for Android
what
depends on ur pc gpu and what exactly u want to do
ai for draw: #1309565672195620935
ai for voice: #1289538710307602554
Hello friends, I haven't used these programs in more than a year and from what I see it is updated,
I had a discord group where I could find many models but it seems to have disappeared and I can't find the one I want (it doesn't work on this server)
Could you help me by providing me with a server where I can find artists? 😦
OOOOOH, MUCHISIMAS GRACIAS
thx bro
huggingface is so weird, are you telling me I can download each file of a dataset one by one, but nobody thought of a download all button?
clone repo should work the same the github does
oh um
nevermind
Wake up, everyone! A new RegalHyperus drum model just released!
Piercing Light (Drum model no. 574)
Retrieval-based-Voice-Conversion
Aight thx
ц
Hey anyone here interested in solona and Ai agents dev Dm me
Realtime Voice Changer
Question
since RVC Disconnected isn't really being maintained anymore, what should I switch to?
Hi
@tacit sierra pwease can u delete this
applio or mainline
Who here eat butt
Hello everyone, can someone make a song with AI for me? I can't do it so if anyone can help I would appreciate it 🙂
just use weights
oh wait, maybe Suno
I only have my cell phone to do it and since I don't understand English it's even more difficult to do the covers
I wanted to make the boyfriend from FNF singing Happier Than Ever by Billie Eilish
I even copied the model, I've been trying for half an hour and it's impossible
🫠
If anyone can help I would be very happy 😄
hi
hi
can someone please send me the rvc download link? im to dumb for github cant find any download thing there
voice converter or realtime voice changer?
can you send me???
i got the link for the realtime voice changer
-realtime
Interaction has expired, use the command again for a new interaction.
you want that?
First link.
this same
ima send to you dm
ok
Nah, no need to send the link in direct message. It's publicly available. 
FOR THE LOVE OF GOD, SOMEONE PLEASE HELP ME 💔💔💔💔💔💔
!howtoask
How To Troubleshoot 
- Don't simply mention your issue, like "
my rvc is not working". - Describe the step you are on, what you're trying to do, the RVC you're using, a screenshot, etc.
- The more context, the better.
- Don't be desperate. You can ping a Helper, but if they ignore, they aren't available/don't know the answer.
- It's okay if you're frustrated, but don't take it into this server.
- Don't DM without prior consent.
- Don't ask for every little instruction. Put your own effort & test things by yourself.
- Don't ask to ask.
- Check if your answer is a Google search away/on our guides website.
I need help from a human 🫠
To get some help about using W-Okada, go to #🔍│help-w-okada. #✦│chat, #🧬│ai-chat and #✨│ai-help aren't where you get help about using W-Okada.
I'm sorry, what? You can do one by yourself.
I'm from Brazil, I don't understand anything about websites, I've been trying all day, but I can't and I don't have a PC 😭
Would that be too much to ask?
Begging anyone to do it for you really shows how lazy you are. 
MY FRIEND, I'VE BEEN TRYING ALL DAY 😭😭😭😭😭

To do AI cover online, search up weights.gg on Google. First of all, you'll need to sign in with your Google account there before you do AI cover.
Can I do it now with the models here?
Voice models from #1175430844685484042 and on Weights are connected. You don't have to reupload the voice model you've downloaded from #1175430844685484042 to Weights.
Have the audio file of your song ready, & let's extract the vocals from it with an audio isolation software.

my isp is gonna hate me after this
i made a program to scrape the emoji images from a site (apple ones)
for the training model
what is "batch size"?
are u making open genmoji 2
Uh so its kinda how much data is looked at per step I think
You should usually double your batch size according to your vRAM, so 6gb = 12 batch. Changing batch size can help your model converge faster, or help avoid overtraining on noisy datasets
Its good to train your model multiple times, on different batch sizes, and compare the tensorboards to find the best performing batch size. Every dataset is unique so there's no magic "best" size
this is wrong
Please inform me my king
VRAM used for training is ~5GB + ~0.5GB/batch
so your formula 'double vram size' is not gonna work
perhaps with checkpointing enabled
I blame @river adder for beating it into my brain
I never said that
Silly
But good to know. I have noticed my PC act slower with higher batch but I guess its not that much
also more is not better
you probably end up using shared memory, which kills performance
The only thing you should keep in mind is to use 8
Well that's what I said in my model master applicstiom and u approved it
I wasn't the one that did...
Yeah I start with 8, then 12 then 4
Oh ok I'm just blaming you for fun at this point
hello it gives me a problem
it says key or something.
Please be specific on which problem you're encountering.
press any key to continue
No, but like the program you're using. The program you're using like W-Okada or an RVC program.
Yeah, that's the RVC-GUI-main I have right now.
-gui
yes
RVC-GUI in 2025? Damn. 
This RVC fork is too old now. It would be better to switch to the newer one instead.

or ilaria rvc for inferencing
I guess I making a weights port
Let's try
-rvc
- How to use RVC Mainline Colab by Cauthess
- AICoverGen Colab Guide by Eddy (Spanish Helper)
- Create a model with RVC disconnected (colab) by Angetyde
this time it gave a different error
Ottoman Empire
Turkiye
Man i just spent 3 hours trying to set up a voice changer and still failed that’s crazy
Hi, why is it that I can't find not one Colab RVC AI page that works? Can someone help me?
or some other method?
RVC GUI is super outdated, Don't follow YouTube tuts, uninstall it
Tell me ur PC GPU and what u want to do
Tell me ur PC GPU, tutorial link, and issue in #🔍│help-w-okada
Huh?
#📰│dev-updates some colabs had problems, I listed here the working ones and the alternatives
I’ll do it when i wake up thank you
Oh thank you
Yw, @ me when u wake up and gn
Yw
Night
I already uninstalled it, but what does PC GPU have to do with it?
If you don't know. Any Python-related program can work with GPU, especially Stable Diffusion, W-Okada and RVC.
Everything, it's the component that processes it, and AI takes a lot of computijg
What's ur PC GPU and what do u wanna do
If you still have no idea what your PC GPU even is, you can open Task Manager, go to Performance tab and look at the left if there's GPU 0 or GPU 1 there.
nothing to do with the subject
What do you mean it has nothing to do with the subject?
Stable Diffusion, W-Okada and RVC are all Python-based program.
they have nothing to do with the topic, I just said that the application has some bugs

I answered to your "but what does PC GPU have to do with it" question. And you think it's unrelated to the topic? What?
I only said that there were mistakes and problems, you gave an answer that is not relevant to the subject.
Nahh what, I'm not gonna argue you with that anyway.
but I didn't change the subject, you did.
Ok, ok dude. If RVC doesn't work, don't blame me. This is like my best trying to tell you to understand.
I already said that RVC did not open and gave a strange error, you said something cpu or something that is not directly related to the subject.
what application did you refer? (which rvc?) not the old tiger18n's rvc gui?
and if you say it has bugs, you should have explained it including how to reproduce it
You people be trying to get me wrong like if you're very confident of yourself, sometimes huh.
Breh. 
at first I downloaded the latest version, then it gave an error, then I installed python 3.8, then it started to nolmal a little bit, suddenly it gave an error, and when it gave an error, I said I would ask.
rvc needs python 3.9 or 3.10, who said it is 3.8?

ffs... latest version of what?
Damn. None of Python programs today use that version of Python. It's so old.
Only 3.9, 3.10 and greater are commonly used as of today.
With Python 3.8, many packages of their latest versions wouldn't work.
there it said Install Python version +3.8 if you have not: it said this so I installed it
rvc
be more specific
Just because they said to install Python version +3.8, doesn't mean you should install Python 3.8 as the main version.
gimme the link that said 3.8

Oh boy. Am I missing something?
This link is in the Preparing the environment section.
as expected
-gui
-gui
I'm sorry, dude, but like you still stuck with the old old fork RVC program that existed before the better ones. And you're trying to get me wrong about my answer that had nothing to do or related with the topic. 
his understanding is also old

like this did not give you any hint that it is outdated as fuck
Yeah, I downloaded it.
in AI landscape trying a 2-year old project is like trying to install Windows 95 and complaining that it does not work

- How to use RVC Mainline Colab by Cauthess
- AICoverGen Colab Guide by Eddy (Spanish Helper)
- Create a model with RVC disconnected (colab) by Angetyde
I remember I gave him the -rvc, but like he didn't even look at this.
Average rvc user
hopefully he should do so rn 
im gay
Good to know.
Oh I got it
but how can I change its value?
I was testing some models, and they have a lot of stuff to setup
like sample rate, precision, pretrain, etc...
hello
Just wondering, there is a RVC for a real time voice changer?
wdym?
it's like asking "is there truck for a car?"
Yea, kinda stupid i know, but usually RVC links are not made to make covers and models?
RVC is for making models and conerting audio files from one voice to another
yea i know, but i seen some tutorial on YT who use a RVCs for real time voice changer
RVC model can be used in a voice changer
there is a link of it? or something?
i didn't mean that, but whatever, my bad
are there any UVR model that can separate backgroudn noise, not music? so for those with sensitive mics that can catch chair creaks, table soft thump, rubbing clothes?
hey, does anyone know how to connect to weights?
You have already asked for this in #🔍│help-w-okada.
Using the Weights bot on Discord?
This is where you connect your Discord account to Weights.
wuts the best settings fr
Can you be specific on which program you're using?
rvc
W-Okada the realtime or RVC the audio changer?
first 1
The first one, I think you're looking for W-Okada. For W-Okada, go to #🔍│help-w-okada.
#✦│chat message @timid musk let's do a 2025 update
Weights by AI HUB 🔥
add me to!
Poyo when they click on a sketchy link for the 9999999th time 
I click them too ngl lol
then litsa bite me
Someone posted my IP but idc
I used W-Okada from Google Colab to troll Blue Archive fans on Discord. Someone just has mistaken Arona voice as Paimon from Genshin Impact. 
Basically the same one from Automaze, the "modified" W-Okada Colab.
A fck I don't think I updated the kaggle version of the original yet.. I mean it should be working rn
should
can somebody who has access to a model that can change the lips of a video help me out with something
For the past two days, I've been training two image models of Blue Archive girls. But right after I posted the image creation to promote my image models, the creation numbers soared up like crazy. 
Hello, i'm trying to train a model with RVC v2 Disconnected on Colab, i followed every instructions and started the training, it's still at 0% at Epoch 93 i only wanted to know if this was normal or not ?
yes
Hey, I was wondering if Applio is still one of the best RVC interfaces?
does anyone know how to get the role Ai members?
Applio is one of the only RVC fork programs recommended to use.
I remember there was a security issue a few months ago, that's why I asked 😄
You talk here until your name turns light blue with a blue star icon, that is where you get the "AI Members" role.
Wait, do you really need like a AI cover for get that role?
that was blaise's account getting hacked, no actual security issue is in applio
No
you need to talk
like how levels in servers work
Ah, sweet. Thank you for clarifying.
oh ok
The "security issue", I think you either mean by an antivirus program that falsely detecting Applio, which can happen to some people. 
You don't need to do AI cover and send it to a mod just to get the role.
Nah, there was a security breach notice, but as VY1 said, someone hacked into Blasise's account.
quetion. if i wanna outpaint downwards do i have to include the woman and child in the prompt?
iiiiiiiii dont think so?
if anything itd probably generate another set of man and woman
great thanks (the model messed it up so bad i gave up on the first try)
What's the recommended Batch Size for a 10GB vram GPU? (for training)
That will also depend on your dataset length.
How much minutes your dataset got?
Sadly only 6 minutes 😦
Hmm, you could try with a batch size of 4.
Alright! thanks
the rule of thumb is either 8 or 4 for most cases
who has a normal Russian female voice?
all the voice models i use with realtime w okada is just robotic.. idk if im doing something wrong but ive been testing the pitch and stuff and cant get it to be less robotic sounding
using stuff from #1175430844685484042 and it doesnt really sound like it
Merhaba
3
Hi, could someone please send me the download link for the application?
replied in #✨│ai-help message
a StarCrafts enjoyer in this economy??
@dusty sun check dms
commands?
with the new 50 series about to release is 5080 16gb vram going to be good enough to game+use a realtime voice changer like okada? (sorry I don't know a lot about how it works) or would a 5090 be better for that.
Of course. Any NVIDIA GeForce RTX 30 series and up should work.
as long as you can fps in-game
otherwise they'll be wraging resources from each other
Anyone have any good voice changers are actually easy to use?
since when the hell i had that?
you have caught the one in #1312246655877124227 or probably in another server
it is more than enough for voice changer, but not a good value for being 16 gb instead of 24 gb
also 5070 Ti is 16 gb, though depend on game spec you want to run
I currently run in to some issues with a 3090 on some games having to drop settings so much or cap FPS. I was kind of hoping that newer AI stuff in the 50 series would help alleviate that issue, but I wasn't sure if 16GB be enough to bring up some of those settings and work smoother, or if a 5090 32gb would be better.
Baiscally just want a smooth experience (as much as possible)
the game may be demanding, try fps cap to 60 and lower the game settings
RTX 3060 or decent 8 gb cards, or 4070 as recommended spec are actually enough to run typical valorant, etc
if you really want to run demanding ones like alan wake, consider using a second gpu/laptop (RTX 4060 recommended) to run voice changer and streaming
guys when I use ai voice changer and I talk with it my friend hear he's voice too when he is talk from my microphone
how to fix that?
use headphones
I use headphone
you mean not cable?
i mean the incoming audio with your friends's voice needs to go to headphones, not speakers
so you mic does not pick it
yeah Im in headphones not speakers
and they hear their voice with the ai changer
like when they hear ther voices they hear it changed
using vocemeeter or something?
you need to figure out your audio routing then
so your incoming audio does not go into the voice changer
and how I can do that?
headset mic -> voice changer
as long as your mic does not pick the audio from the headphones there should be no issues
when I don't use the voice changer my friend dont hear his voice
its something in the voice changer
and when I hear myself in the voice changer app and my friend talk his voice change too and we hear it in the voice changer app
not in discord
so the sounds that coming from my computer when I use the voice changer they go there too and changed
so how I can make the voice changer just get the sound from my mic only not from the other sounds
closed-ear headphones that don't leak are ideal, but you should try lowering your friends' volume and increase the threshold
ayo that's a gif I can hear!
So I was creating a silly song in Suno, but for some reason I cannot generate it. But if I try replace every word that says “Fortnite” with something that sounds similar such as “fortnight”, it lets me generate the audio anyways. How cool is that?
hello, this is a bit of stupid question, but on the kaggle tutorial
Use batch sizes of 4 for short datasets, 8 for minimum length datasets (5-25 minutes), 12-16 for longer datasets```
So does this mean if i want batch size 8 for 5-25 minutes length dataset like on the guide, i set it as 4 on the rvc? as it's per GPU?
Dead ass chat
timezones exist
@polar flax
💀
damn, two scam links in this economy? that's crazy
it's kind of crazy that some people still click this links and get hacked lol
Ikr lol
anyone give me rvc link
- How to use RVC Mainline Colab by Cauthess
- AICoverGen Colab Guide by Eddy (Spanish Helper)
- Create a model with RVC disconnected (colab) by Angetyde
What's ur PC GPU and what u tryna do
ONLY english is allowed
Suno’s guidelines about using trademarked names is kind of stupid TBH.
why is my voice ai keeps crashing?
Does anyone have a voice model where it swaps between types of voices like as seen in this video at around 0:37? https://www.youtube.com/watch?v=2Uy_2gxLTuo
or is that even possible?
are you talking about https://voice.ai ?
Download now free real time voice changer for pc. Modify your voice, select one from the Voice Universe or clone any voice you want
because that one sucks and we don't use it since it uses your PC power for distributed training, we use wokada
-# I'm guessing you want a realtime voice changer for calls, so wokada
tell ur pc gpu in #🔍│help-w-okada
It has probably been discussed to death, but I'm overwhelmed with information and don't feel any closer to understanding how to tackle the problem.
I have a recording in English, acted with voice A. I want it to be acted out in a different (minority) language X with the same voice.
A short recording, few lines, couple words each. I found TTS/RVC models of the voice A, so I don't need to train. I have the translation to X. But how do I approach making the voice speak X? I can record myself, but as I understood, it will still have English accent (and I can't act for shit, but that's another problem).
just use a tts to create an audio file in language X and put it through rvc to change the voice
Some Italian TTS produce decent result, but that approach would strip the acting intonations of the original and make the delivery flat, wouldn't it? Wouldn't also the English accent of the original voice still be heard in the output?
AI HUB Docs


