How to Change Your Voice in Realtime Without a GPU or Powerful PC (For FREE)
ฝัง
- เผยแพร่เมื่อ 16 มิ.ย. 2024
- Today, you will learn how to use W-Okada's free AI Realtime Voice Changer without a GPU! Transform your voice into your favorite TH-camr, VTuber, Anime Character, & more, all in Google Colab
Colab link here:
github.com/w-okada/voice-chan...
Original Tutorial Video:
• How to Sound Like an A...
Here is a video on how to get it to work with Discord and other apps:
• How to Change Your Voi...
Join the AI Hub Discord:
/ discord
Discord Full? Check out our fix video:
• Where to Find AI Voice...
Check out our new video on how to train your own model!
• Video
Search the most complete list of AI Tools, also available in 中文, español, 日本語:
ai-search.io/
DISCLAIMER:
Please do not use these models for malicious, harmful, or deceitful things. Please use them to have fun and experience this new technological age.
~~~~~~~~~~~~Timecodes~~~~~~~~~~~~
Intro - 0:00
Loading up Colab - 0:31
Ngrok - 3:28
LocalTunnel (Ngrok Alternative) - 4:47
Using the Software - 5:06
Testing Model - 6:36
Uploading Custom Models 1 - 7:04
What to do if yours Crashes - 7:34
Ngrok Working - 8:47
Uploading Custom Models 2 - 9:29
Testing Custom Model - 10:17
How to delete - 11:19
Can you use on Mobile? - 12:22
Wrapping up - 13:20
~~~~~~~~~~~~Timecodes~~~~~~~~~~~~
Here's our equipment, in case you're wondering:
GPU: RTX 4080 amzn.to/3OCOJ8e
Secondary GPU: GTX 1080 (too old, would not recommend)
Mic: Shure SM7B amzn.to/3DErjt1
Secondary mic: Maono PD400x amzn.to/3Klhwvu
CPU: i9 11900K amzn.to/3KmYs0b
If you found this helpful, consider supporting me here. Hopefully I can turn this from a side-hustle into a full-time thing!
ko-fi.com/aisearch - วิทยาศาสตร์และเทคโนโลยี
how do you go about updating the version running locally? do i just drop in the new version folder and overwrite existing files or should i make a new folder and port my uploaded models? just curious what you think.
THE AI OVERLORD HAS CAME BACK AGAIN WITH A GREAT TUTORIAL VIDEO!
Google has banned peeps from doing this kinda stuff of late, you may be able to connect thru, but get kicked outta the system or disconnected every few minutes!
Is there a way to bypass this? maybe another platform?
Could you do this on runpod?
I've moded the code to bypass it and it runs 2-4 hrs straight... But unfortunately Im not giving it for free ☠️
How do you fix this , I have AMD and didnt download it because it said it wasn't compatible
Does it work now or is it still incompatible with AMD ???
@@nilakshidharmarathne5351 I was doing in a modded version of this, it worked smoothly there
You have no idea how much you’re helping creators from around the world. God bless you❤
Idk about creators but I am so gonna harass my friends with this lol
@@stumbling_ mee too bruh a little bit of trolling
you can use android emulator on your pc if you want. you can use on mobile, but you still need a pc to get this works.
what do you need is audio jack 3.5mm to y splitter (mic + earphone) and audio jack 3.5mm (please choose carefully that there are two type of y splitter, DON'T take the dual 3.5mm female one.)
connect the y splitter to the phone and connect the audio jack from your audio interface to the y splitter (mic). And you're good to go.
hope this help.
Note: you still need a laptop or pc.
Could ya help me?
@@Anime_IT. i will do my best, what can i help?
What for?
Collab works on Android but i'm not entirely sure about this notebook.
Source: I'm an AI researcher that used Collab for machine learning. I'm now self hosting Jupyter Notebooks.
I learn that the voice would get choppy if used for a while. It Easily fixed by just refreshing the page. I am not sure about others but it work for me.
Thanks for the tip!
I wonder whether we see some hardware solutions for stuff like this at some point? Something like a Roland VT-4 Vocal Transformer, but where you add voice models via USB directly to the device and then let it do the processing in real time while you can tweak it on the device itself.
That would be cool!
any plans on making a vid on how to install this locally? since colab isnt working now
that was surprisingly easy, tysm :))
You're welcome 😊
How in the world did you get this to work? A lot of people are having issues
OMG BRO YOU HELPED ME SO MUCH
Google colab works on any mobile browser.. Just turn on the desktop site mode.. Pretty cool..
Nice, thanks for sharing!
getting this issue
" Frequent errors occur. Please check if the model of the framework being targeted is loaded."
when loading voices or trying to select a vooice... :( WTF
same
People 1 month ago didn't have this issue so I think Google screwed something up very recently.
did you found solution ?
So it turns out that due to Google restricting gpu usage, nobody can run this using colab without paying for extra time. Everything seems to point back to the idea that running it locally is best (and that you have no choice but to buy a gaming pc).
@@thewellknownemployee8512 no, i just pay like $5 now and it works fine, or i think its like $9
better than installing something locally. as I never like to download any programs, i hate taking that risk.
It doesn't work to me as well but the alternative local tunnel does not show on the link you provided. Is there something I can do about it?
I have this same problem
Hey man, god bless u have a nice videos, super good i love it
Thanks!
Hey can i ask u how can i connect the voice to discord?@@theAIsearch
This was so useful, thanks for the information and helping!
Unfortunally I'm experiencing some problems which are just bothering, you see after getting it all done and get on ngrok, after some minutes (normally 10) it just crashes of using it or not, every time. Is there a way to fix this or I'll need to do the crashing process every time?
Colab has now banned this for their free plan, so you'll need to install it locally or pay for Colab
@@theAIsearch Oh, I see, well thx for telling!
Well worth a subscription
Frequent errors occur. Please check if the model of the framework being targeted is loaded.
Nice video btw ❤
I have a question for realtime best gpu is 4090 or 4090Ti?
ti tends to be better for ai tasks
It works perfectly, but after a few minutes it stops working in collab and I need to redo the process. Thanks
3-6 months ago, they implemented a limit to how much GPU or Resource you can access. I was there before there was no limit and while I was abusing the free ai image generation on Google Collab, I realized that they added that limit and they want me to accept their premium.
Im an AI / ML engineers should have pick this as my thesis
I don’t have a third option it’s only clone repository and install independents aswell as
Start server using Ngrok which Ngrok did not work
Can you make a video where you imitate the voice of a famous actor? I don't know if it's worth paying for the collab so that in the end the voice doesn't sound good.
Google collab is giving me an error saying that I tried to run code that is not allowed in the free version. I am just going to run this on my Linux laptop. It is not that bad, Nvidia GTX 960. The windows version does not work on my windows laptop lol, but it works on that Linux laptop on a virtual machine but because of the virtual machine it does not have access to the GPU yet it still kind of works but it is unusable. So yeah I will just take that collab code and edit it so that it does the right things on my Linux system and then run it on my laptop without ngrok.
update--- I got it to work but it only runs for not so long.
I'm having an issue with my access to Google Colab. I'm unable to log in due to an access restriction.
Can you please help me resolve this issue?
I appreciate your help in advance.
Hey can you try helping with the Kaggle version?
Yeah I tried it on mobile but it always crashes or my big fingers deleted all the codes so my recommendation is use laptop because its running colab so crapy one can also handle , like I use old 32 bit processor laptop and its work so use laptop in my recommendation
everything worked out fine but this message keeps popping up and shutting down the site after a few mins "Your runtime has been disconnected due to executing code that is disallowed in our free-of-charge tier." i followed the steps exactly and there is also no option to use local tunnel or connect to google drive
This is happening to me too. I would have to reinstall it everytime, I wanted it to work again. This wouldn't be that much of a problem, if I could speak for like an hour, but it literally shuts off in minutes!
Same here too
i have the same issue
is this malware?
Google ban it give up
I tried this in 2024 with no success. Not only had colab banned guis, but they also made it so the ngrok link gets disabled before the model gets loaded. That is pretty annoying and deserves an updated video of some kind. Even the discord help link is missing!
Thanks for uploading these baby like step by step videos for a braindead person like me 😂
Happy to help!
@@theAIsearch for some reason both of these methods dont detecte any input or output devices i have no clue what the error is
doesnt work 1. i dont even have option google drive, with ngrok it doesn work and i dont have LocalTunnel option :|
Hey man really nice video can i convert audio files with this??/
try this th-cam.com/video/ixB9oalT3cQ/w-d-xo.html
im sorry the google drive option seems to have been removed what do i do now??
Is there a way to not hear myself from the voice changer program like if there's some sort of button I can press?
i dont want hear the voice of the voicechanger when im talking, im in laptop how i can fix it?
google colab is not free anymore they say 0 compute units
google colab is a free of charge
I dont know why but when I open that I dont have google drive to run
no option to connect to google drive ?
Can you make a video on how to hook the voice changer to your phone and console please
why cant i use to google drive option, i mean its just not there
how do u save it too google drive there no link
Mine is crashing after like 7 mins or something its not crash but it saids rundown time ended or something by collab, Is collab being funny? :d
Colab has now banned this for their free plan, so you'll need to install it locally or pay for Colab
updated tutorial please???
For me it loads up for a second and then an error pops up with two buttons saying initialise and reinitialise 😢
What do you mean by installing locally?
Is this able to be used with the web version of Discord?
is there a way to record the output of the generated voice? for voiceovers
DID U FIND OUT
could you make an updated version?
I recall google collab tends to have timeout, like if your inactive for 5 minutes, then you will be disconnected and you will need to begin to whole process again, is this a thing now? host?
colab is disconnecting GUIs like okada now. its best if u can install this locally
I am trying to click the characters and it isnt working lol
There is still a delay or lag when using it on Mac
Im curious about that marin kitagawa voice model
the one now doesn't have the drive part which is okay for one time use but is annoying for multiple times using it
what GPU do recommend for this software? , i was thinking of getting a Radeon RX 7900 XT but im guessing the intro that wont work
nvidia cuda gpus are recommended for most AI tools
hi, how do i reopen and use the voice changer after i closed it? i have to redownload everytime, pls make a vid on it i beg u plss
Can you use it with your headphones microphone and not an actual microphone?
You can as long as the quality is good enough. Make sure there's minimal noise or echo
1:33 the optional connect to google drive part is gone
Yeah 😢
ur natural voice kinda sounds like mr beast
Also the thing called "the optional thing 1:43 I dont see it in my link
maybe because you are already signed into google on your browser
Everytime I open Okada, it asks me to “Initialize”, I tries “Reload without initialize” but it won’t work. The program only run if I “Initialize”, which cause me to lose all settings. All firewalls, and antivirus are off.
sadly the only way to have the website up is to buy the compute unit colab thing. the other options like the google drive thing and the local tunnel doesnt pop up for me so i think its broken or smth.
colab is disconnecting GUIs like okada now. its best if u can install this locally
i have been trying to use this voice changer for awhile now every time i finally get to where you can use it after a while it says you new to connect to a runtime so i do and nothing happens can anyone help me figure this out?
You'll get disconnected after a while with the free plan now. You need to have a paid colab plan or install this locally to run it.
Nice name btw
And I Am the second comment I love your videos bro
Looks like you are! Thank you!
For some reason starting yesterday, the google colab end runtime cause of some restriction. Not sure how and why but it seem it keep stopping runtime with the pop up after few minute the server is up.
in the output and input theres no options for me just "none" :(
Please I need help, I'm trying to find more natural deep voices on the discord models but I can only find a few like Corpse and Markiplier but I want more, just a natrual sounding one. Can someone help?
Can this download to phone?
I Confirm you can it use this in your phone I'm running this to 2gb 32gb phone
do you need to do clone repository whenever you use it? also, the runtime crashes always what should I do with this
the colab free plan will disconnect this now. see my other video on how to install okada to your pc
i downloaded this and its not picking up my gpu and its nit changing my voice
am I the first one to comment? And thanks alot this is helpfull
Looks like you are! Thank you!
hi im having trouble it keeps saying pipeline not initialized when i press start on the voice changer and no tutorial is there i even installed python
When a voice changer is used, does it have to have a stable signal?
i am getting this error please help me ERROR: pip's dependency resolver does not currently take into account all the packages that are installed. This behaviour is the source of the following dependency conflicts.
lida 0.0.10 requires kaleido, which is not installed.
torchdata 0.7.0 requires torch==2.1.0, but you have torch 2.0.1 which is incompatible.
torchtext 0.16.0 requires torch==2.1.0, but you have torch 2.0.1 which is incompatible.
torchvision 0.16.0+cu121 requires torch==2.1.0, but you have torch 2.0.1 which is incompatible.
Mine too, does this effect anything?
Answer this question man
when I open the link for google colab just the first option its available and its not working :( im sad I really wanted to try
the localtunnel option does not exist anymore....
Thank you so much.
You're welcome!
i have a question: why i cant choose crepe ;-;
What about windows 10?
google stopped allowing GUI apps , will be disconnected within 10 -15min
no sound comes out, everything opens normally, I'm using VBCable, I installed the compatible versions of pyton, pytorch, cuda... and there's still no sound coming out
same
why is it when adding new models, it just doesnt show up after I upload it
Can I upload like a 30 minute audio in arabic or english of speaking onto this and download it fully in a completely different voice?
this is better th-cam.com/video/ixB9oalT3cQ/w-d-xo.html
getting 2 mutch runtime how 2 fix
When i play the (Clone repository and install dependencies)
I get an error called (ERROR: pip's dependency resolver does not currently take into account all the packages that are installed. This behaviour is the source of the following dependency conflicts.
lida 0.0.10 requires kaleido, which is not installed.)
Can u please post a video on how to fix it?
its fine it will work
Ignore it ;D
How can i supress other sounds like my breathing, cuz she does weird noises when i dont talk
Also do i need to do the first step "Clone repository and install dependencies" everytime i wanna use it or its saved, like when i quit the site or power off the pc ?
you can lower your input volume.
i cant select input or output , it just doesnt appear. how to fix this
dude when i go in the link and preview , it just keeps loading and never shows up
Excuse me, why is there no 2nd option the localtunnel when I try to find it?
yo, could u make an updated tutorial because some of the buttons have changed i dont wanna explain it to my friend
I second this
it says not enough computing unit after a few mins that i start it, is there any fix?
there is no option called connect to google drive
guys, it's me or the the github it's gone ? it shows me " Unable to render code block " ...
It does not work it says module error modulenotfound at the 2nd step
for some reason i do not have the option to connect to google drive. It is just not there. Any assumptions of why?
ERROR: pip's dependency resolver does not currently take into account all the packages that are installed. This behaviour is the source of the following dependency conflicts.
chex 0.1.86 requires numpy>=1.24.1, but you have numpy 1.23.5 which is incompatible.
pandas-stubs 2.0.3.230814 requires numpy>=1.25.0; python_version >= "3.9", but you have numpy 1.23.5 which is incompatible.
torchtext 0.18.0 requires torch>=2.3.0, but you have torch 2.0.1 which is incompatible.
torchvision 0.18.0+cu121 requires torch==2.3.0, but you have torch 2.0.1 which is incompatible.
how to fix that?
dude. Run connect to google drive isn`t here anymore
can you mkae a updated version plz!!!!
good, but i cant choose gpu, only cpu (im on laptop windows)
OMG FINALLY
bro i downloaded the driver but after downloading the driver speakers not working ! no voice . help me pls !