#🧬│ai-chat
1 messages · Page 333 of 1
Nvm, it is individual preference. You like light but I like dark. Because we both have different taste, different choices, different reasons to use these modes. That's why developers provide both.
👋
both gone.
oh nvm kalo is still there
just extremely inactive
does anyone know if running wokada in the background would trigger valorant anti cheat? thinking since its running in a cmd
I think I was the first one to make the Kalomazing mode joke back when previous AI Hub server. 


in voice models i can't see anything
that might be an internet/idscord issue
you have the perms to see them
everybody
be sure u have cleared ur research in that channel and to both refresh it
You actually run W-Okada in the background while Valorant is running?

havent tried it yet too scared for that anti cheat spent way too much on skins to get that ban
just dont risk it then, the latina e-girl voice can't be worth risking your account getting banned
yeah but the fun you could have in swift play just seems too good
the dilemma of an valorant player
true probably lol some valorant players are rancid for egirls
i say this based on one guy i edited his clips for a long ass time ago

i should probably do that again out of fucking boredom but i am so washed and i dont see any point
i rarely do any of that cool shit now lol
lmaoo that depends on which server youre on its either simps for egirls or just straight misogynics
i wasnt even in a server he was just a guy me and my friend talked to for years
he was rancid ever since we met him anyway
yea checks out, most players are a special kind of degenerate
Ayo? @worn cape level 3 !!! 
Gooning while gaming. 
nooo dudeee 
resorted to using it for trolling in vrchat, so far so good
although i do really need some extra ram im currently frying mine
Her name is Lucy. 
He is gone ?
No! My YouTube channel / account has been terminated by YouTube! NO!
lol..?
what did you even do to get it deleted
From here yes
Ya he was a great helper
from ai covers?
Yeah, I mean, I saw one of my videos were removed since of copyright
3 videos actually
So now I actually got 3 strikes at that was it
yea thats why no one does ai covers anymore
- the hype for them died
Ohh.. so for personal reasons
i rarely see ai covers now, before there were whole channels just for them
Well byee poopmaster
Yeah, and now, I lost 173 hard-worked subscribers
Ayo? @bold rampart level 5 !!! 
And lost 190 videos that I made and posted on that channel!
are u gonna remake the channel with again other ai covers till it get taken down
Because it is now very easy to do that. But it was quite difficult then.
Not really
when ai covers did a big boom was when there was rvc
No, but I really like making AI covers and posting them on YouTube, but now, I'm afraid I'll get copyright "infringement", thing, I think
so nothing really changed much
rvc is still the same since a year
in terms of quality and making a model
welp, yea that’s what gomma happen
unless u mean when there was so vits svc, but i think the peak was when there was rvc
I've started doing AI sloppery on YouTube last year. My channel still hasn't yet gone, I only let the copyright claim issues die at time, not much to do it. 
I mean, now, there are many websites and apps for this. Even a person who don't know anything about RVC can Use apps like Weights.gg
i joined at basically the peak july 2023 when ai covers where everywhere and it was rvc
the main thing isn’t being hard tho
the reason why the hype died is people got bored

Yes, you are right too
and also bc copyright ofc
So, what should I do? I have recently put my Patreon page on there, so that I can earn money, but, sigh, all of those were in vain
When it became common. It die.
At the end of September 2023, I took a break from internet for a week because of a crazy beef with a kid, and then everything fell off right my face. 
And also when there’s no updates to it
e
in november 2023 i deleted my account and didnt bother contacting anyone from here again for no reason
welp, either make other type of content, or retry on youtube, or split the cover in multiple under a minute part to post them on tiktok (doing like that, copyright isn’t affected)
RVC and Stable were unusable on Google Colab for good. Got banned off from a Discord server for the problem I didn't cause. 
It has been a year since I went missing in action. Now that I think it's time to go back to the business so everyone can love me again as before. 
they are still both banned btw
just now we use encryption
"Make other type of content", like what?
What if I use Ai covers on social media apps like Instagram or X. Can I earn or I will have copyrights ©️
Like what u like, idk
U could do tuts, gaming, other shit
I used to do fortnite glitches + showcasing unreleased/dev shit before rvc
Now that I can only use Google Colab for demucs. For RVC and Stable, I can only do them local because some weirdos from old AI Hub told me to do so. 
I'm not really the type of person that can switch on the camera, game with camera for like several hours, and post it on YouTube. That's never gonna happen. Anyways I really don't have any ideas what to do...
Im still known from that and still get comments on how to gt those patched ways 😭
I was for a long time in the fortnite glitching community
now for example I made yt tuts for my own AI projects
obfuscation even
btw, one day could be that gets patched
Just don't give your too much on YouTube and AI things, bro. 
Create tech related content, like how the processor of a computer works, different types of keyboards which one is best membrane or mechanical etc etc..
yeah they'll probably just make bard detect it or some shit lmao
@covert lake
hey chat, this is probably the wrong channel to ask in, but what's the best AI for youtube short videos generator, i was thinking of crayo.ai, but i'm not sure
it seems colab uses a diff type of detection for deepfakes, for example with multiple layers of obfuscation & encryption, facefusion is still unstable on google colab
Like risk of stopping mid run
For some could be 2 hours, some 3, some half an hour 
Which is why I mainly suggest people to use lightning.ai that has no restrictions and gives free gpu
What if I use Ai covers on social media apps like Instagram or X. Can I earn or I will have copyrights ©️
Asking 3rd time
I dunno
I don’t use instagram
idk why you would be trying to make money off of ai covers
Free GPU or not, if I have money and their service is good, I'd buy them. 
Me too. Just asking. What if I do in future
I use X/Twitter but i think that has a copyright system too soo
for example, when i used to do mvs leaks on X, they used to copyright them down sometimes
I don't make money off from doing AI generated sloppery. All I do for AI is funny. 
Its prolly better to make money another way
Yeah, I could, but I really don't have the knowledge to do a 3D animation for explaining how tech works nor have the processing power. Even though I own a laptop that has a Intel CORE i7 12th Gen 12650H CPU and a NVIDIA GeForce RTX 3050 Ti GPU and a 144Hz display with 2 8GB High Performance DDR5 RAM
Well, ig u gotta find something else u like
btw my bad i didn’t see it
Like how I draw art straight with my hands, let people request or commission me with their money. 
Im thinking of either making tuts or doing something unique
Hei
wouldn’t that be enough for blender
Can I share my music here?
Just learn and start. Don't make excuses if you want to earn money.
i remember using it on my i3, i mean just opening a 3d model
#1159290752195633273 as yt link
Okkk
Yeah, I think I should
Thx
Yw
Not a problem
I could make a simple donut in Blender, or a house with furniture, that's all I know from my knowledge
Don't use 3d animations. Just explain technical terms and use photos. Or download the copyright free images and use them.
I mean depends on the style he wants
u needa be unique to emerge from the other thousands of content creators
Yeah, I think I will, and btw, I recently subscribed to the new Hailuo AI Video Generator to the Unlimited plan, so maybe I could try using those and then merging all the generated videos about tech into one
Review whatever you have, borrow stuff from your friends or relatives and Review it on your tech channel. Also tell people about new technologies and gadgets
BTW find open source projects and use them for free as you have good specs in your pc. Don't waste your money on subscriptions
Yeah, I think I should use Pinokio, which is the AI local installer thing
Got it . All the best
That's what I'm planning to do. But I also shared my ideas with you. Good luck @bold rampart
Alright, to you to
That's what I do. 
AI generated tracks, especially from Rizz Records and King Willonius, are kinda copyrighted and also being listed on Shazam. 
if its on shazam it was likely published through something like distrokid
literally anyone can do it with anything
alo

ello
hiii
Guess what was removed. Sorry for the bad camera quality.

Could you export it to mp4? The video won't play on my device.
you dont have the hevc extensions or some shit or ur on linux lol, it should be playing
Alright, gimme a minute
you just exported it as the same codec so its still not gonna play for him
Seems like clothes
It's not clothes, it's blankets
I had to download it and play it on VLC, and it worked. So is it some video overlay blend test right? 
No, I just made motion extraction from the video
From this tutorial: https://www.youtube.com/watch?v=woj4vfMLpao&t=60s&ab_channel=LazyPosy
How to achieve the Motion extraction effect in Davinci Resolve and Adobe Premiere Pro. And chair update... Sorry about that.
Music: https://posy.bandcamp.com/
Support my projects on Patreon: https://patreon.com/posy
Main channel: @PosyMusic
Posy on Spotify: https://open.spotify.com/artist/3zkrmBzisZKngkrd6gXLHg
Or on Apple Music: https://mus...
I see. Maybe next time just export it as h.264 for preview, not h.265 or HEVC. 
Worked fine for my phone
Ok, I'll try again, and upload the file again
Try this time
It worked now. 
oh
Hmm...
Hello
hai
ayo
huh
Hi.
is that yi long ma?
I think so.
Pest tête cabo
Hi
hi
Has AICoverGen really not been updated in like a year?
Ayo? @mighty idol level 1 !!! 
Or is there like a new version of it that I am unaware
Can anyone help
embed misfire even
what happened?
Well I'm training a model I wanna resume but while training I unticked save_latest_ckpt and the step number is different from 2333333 and I can't remember it
After switching account
what are you using?
RVC DISCONNECTED
1 sec lemme compare to mine
?
nothing just checking if step number is stored in log
I can't find step number
in model's log file it doesnt say anything like "saving model and optimizer state at epoch 5 ./log\modelname\G_130.pth" or something?
or checking tensorboard charts, they show step counts
why doesnt the wokada voice changer download link work does anyone has the link that works?
When I try to resume it keeps starting from 1.
wdym
can you send the link?
Ayo? @mint wave level 1 !!! 
k ty
yoo can yall please help me do you know if theres an ai that can make a song using a reference track but it keeps the same beat and notes of the vocals so it sounds exactly like the original song except with different lyrics
Ayo? @lost sun level 4 !!! 
@chilly lake nah it keeps restarting from 1
sure
join vc
Ayo? @golden ivy level 1 !!! 
oh wow
Ayo? @vital snow level 10 !!! 
fr?
cya
i never heard myself sing with it
you train it urself?
lol

Who know good russian model?
Man, it's 11:54 PM here, I've played from 6:00 AM in the morning without stopping man, except for eating or drinking
I am a NLP engineer
Hmuuu
noice model
Your name is a car brand. KIA 😅
Hello
Someone here in to AI-Model creation business?
Ayo? @old seal level 2 !!! 
if u are lookin for someone to make it for u: #1159289738314919936 or #1191429836321849435
hi again, im new and even more confused than i was yesterday.. i want to make ai song covers by using my friends voice but i dont know how to train a voice model and how to replace the singer with my friends voice model
I remember talking with youu yesterday and you used an old yt tut,
Your pc was weak so u had to do it cloud
yeah you tried to help and suggested ilaria but i couldnt even upload the .pth and .index so
ye bc u didn't train the index
i also replied a fix for u about this #✨│ai-help message
reminder that the index is an important file that contains the accent of the voice
U could try to see how it sounds like without the index
Next time you train, don't follow a yt tut
sry if my questions sound dumb but this is all rly confusing to me.. how can i test what it sounds like?
Firstly, you need to read the Ilaria RVC Zero Guide
Table Of Contents Introduction (with website link) Model Loader (Download & Upload) Inference (use RVC AI Voice Models) Ilaria TTS Settings (Inference) Vocal Separator (UVR) Troubleshooting “No GPU is currently available for you after 60 seconds” “GPU task aborted” “You have exceeded your GPU ...
You need to go to the huggingface space, go to the model loader tab and upload the .pth
you then need to get ANY random added index, without any u cant upload it
i did that but i did something wrong and i dont even know if the model was trained correctly #✨│ai-help message
lemme help u there then
hi
will the new m4 lineup chips be able to handle rvc and other ai models (for training etc)??
😂
I don't even think rvc has support for mac training
and i even doubt more that mac could train
applio can do MPS backend
no wrinkles (person i was replying to got banned because he has literally no wrinkles on his brain)
@shrewd iris no need to spam or send random insults in chat
For training too?
no wrinkles
hej jak sie masz
if you wanna melt you mac, sure
-rt
This interaction has expired, use the command
/guidesif you wish to see it again.
Did anyone successfully train or mac?
Ayo? @shrewd iris level 1 !!! 
oh i can't
nie mozesz tak mowic
What are y'all doing for Halloween tomorrow?
marketing BS "40-core GPU" aside, i dont see why that may be an issue
It sounds like you’re excited about a Diddy party! Here are some ideas to make it memorable:
Theme & Decor
Black & White Elegance: Use a classy black and white color scheme, reminiscent of Diddy’s style.
Balloon Arches: Create a dramatic entrance with a balloon arch.
VIP Lounge Area: Set up a section with plush seating to mimic a VIP club.
Music
Diddy Playlist: Curate a playlist of his biggest hits, including tracks from Puff Daddy, Diddy Dirty Money, and collaborations.
Live DJ: If possible, hire a DJ to keep the energy high with hip-hop and R&B.
Food & Drinks
Signature Cocktails: Create drinks inspired by his albums or songs. For example, a “Ciroc Mule” using Ciroc vodka.
Appetizers: Serve upscale finger foods—think sliders, gourmet popcorn, and fancy cheese boards.
Activities
Dance Off: Host a dance competition featuring iconic Diddy dance moves.
Karaoke: Set up a karaoke station for guests to sing their favorite Diddy songs.
Dress Code
Glamorous Attire: Encourage guests to dress in their best party outfits, reflecting the glitz and glamour of Diddy’s style.
Ayo? @fallow tusk level 1 !!! 
well, not too fast, not really designed for number crunching for a long period of time, but why not
what would be the minimum for mac training then?
@forest quarry did u ever try
MPS (Metal Performance Shaders) Pytorch is the only requirement
it may offload some unsupported functionality to CPU
Two reasons: 1) MPS has limited support in pytorch (like complex numbers didn't work properly iirc), some ops may be not implemented 2) MPS is about GPU, it does not utilize NPU so it may be significantly slower
hi
as far as I know, M1-M3 werent much faster than regular CPU
but those are not training values, just inference
Would u ever suggest anyone to do mac local training rather than just cloud? if its just like a cpu it's not even worth it
overpriced asf damn
nah
ahh i love when nobody gives a fuck about my messages
So basically just change the lyrics of a song, not sure if Suno or Udio can do that
But for RVC, u'd have to sing the modified lyrics and run the with the program
just wait, someone will anwser to ur messages in any moment
hello does anyone know any good free ai tools i could use to create concept thumbnails for youtube?
hi

How to make ai character voice
wtf did i even read.
what's ur pc gpu
Not high enough rank most likely
Ayo? @elder willow level 1 !!! 
Which program is the best for the ai music?
?
This is the name of this program ?
Ohhh
Okay
Bruh, anybody here is familiar with the RVC-Project WebUI?
github: https://github.com/RVC-Project/Retrieval-based-Voice-Conversion-WebUI?tab=readme-ov-file
I was trying to follow the tutorial on: https://github.com/w-okada/voice-changer/blob/master/tutorials/tutorial_rvc_en_latest.md
But am i too dumb?
The tutorial said i should start the start_http.bat, but i couldn't find that file, so i went with the next best thing... go-web.bat... but uhm...
Terminal history:
.\go-web.bat
F:\VoiceAI\Retrieval-based-Voice-Conversion-WebUI>runtime\python.exe infer-web.py --pycmd runtime\python.exe --port 7897
🇧🇷 O sistema não pode encontrar o caminho especificado.
🇺🇸 The system cannot find the path specified.
F:\VoiceAI\Retrieval-based-Voice-Conversion-WebUI>pause
🇧🇷 Pressione qualquer tecla para continuar. . .
🇺🇸 Press any key to continue. . .
Am i being dumb?
ask a mod for the "trusted members" role
i really am just here for the tech support and the voice models
you can be here for more things besides support
ik
No only level 5 gets you embed perms
i had level 5 but i couldn't send images in https://discord.com/channels/1159260121998827560/1159290096458149938 or in general channel. iirc
i might just give up and use a speech to text to speech instead if i can't get this fixed...
Ayo? @pulsar forum level 1 !!! 
The channel link shows no access for me
how strange
@earnest dragon @pulsar forum aren't a model maker yet ig
same 4 me
And I'm level 36 must only be for model makers
idk, but just in case you must request for the trusted members role
Yeah never realesed a public rvc model
u should if you want access to https://discord.com/channels/1159260121998827560/1159290096458149938
But what is there really that that channel is good for
good for what
That's what I'm asking what is the channel good for
How do i fix this error message?
what are you using?
illaria rvc
Aight thanks! Finally can have a working voice changer! Let's goooooo!
wait, how do you download it?
-rt first link
This interaction has expired, use the command
/guidesif you wish to see it again.
Tons of info about making models and such. And you don't have to interact with people who have no idea what they are talking about
How to make a video
-rt
This interaction has expired, use the command
/guidesif you wish to see it again.

Ayo? @fringe crater level 1 !!! 
Sup! I have isolated guitar track, and I need advice for AI's that can help me make a tab from that track
hi
weird question but i need to ask. Anybody knows if there's an ai that can reduce noise artifact from a recording? not for singing, just a speech recording
rx's dialogue isolate can help
hey if anyone here is good at video editing dm me. i need a video of my discord server and ill pay 5-15 dollars depending how good it is dm me.
@supple crypt this is a longshot but i recently found your vids, i love your voice and would like to get your thoughts on some models i've made! 
Oh? Are they posted here?
not public! i can dm if you're interested! im fangirling so hard 
I dont mind xD
How do I use the sonic voice changer how do I use it I'm new in this server and on mobile
Mobile you can't
Because no gpu and stuff and also virtual audio cables don't work on mobile
I don't know seems like no
not for realtime voice changer, but u can use the voice models for online RVC, such as
https://docs.google.com/document/d/1YbXcLFPaGjhOdG5NFkK3QrucCEpHZBwFUxkeMO8aB18/edit?tab=t.0#heading=h.mfxmgfqiaevu
Table Of Contents Introduction (with website link) Model Loader (Download & Upload) Inference (use RVC AI Voice Models) Ilaria TTS Settings (Inference) Vocal Separator (UVR) Troubleshooting “No GPU is currently available for you after 60 seconds” “GPU task aborted” “You have exceeded your GPU ...
Yep.
it may seem less friendly to mobile users, but despite being unable to use voice changer some of them may have made AI covers using some online solutions
Wha
.fm
Say My Name
Destiny's Child • The Writing's on the Wall
Don't Tell 'Em
Jeremih • Late Nights: The Album
.fm
Kiss Kiss (feat. T-Pain)
Chris Brown • Exclusive (Expanded Edition)
yea the channel is for model maker role only
Erm actually this is the general chat 🤓☝🏼
the ≠ this
“Cause you have no idea on that discussion” 🤓☝🏼
Ayo? @small cosmos level 2 !!! 
Erm well actually I do 🤓☝🏼
My dick so fucking hard right now I’m stroking tf outta my dick right now finna bust my nut everywhere
more like trolls, random teens and ragebait ppl over here, some of them even sound like llm bot
Chatbot mfs. 
damn you got it boss
Is it harvest
I want to post a weights model on here
since I can't train locally on my potato pc
@unique pawn
turned out pretty good
removed due to possible copyright issue
the model?
the dataset I mean
oh ok
@unique pawn ill dm you it
I posted it in https://discord.com/channels/1159260121998827560/1159514067187277865 so hopefully someone will see it and make me model maker
wait nvm
ill tell you when
Ping me when I get model maker so I know
Is okada the best ai voicechanger rn?
is it free?
Yup on github, open source
You mean, it's better than the voice.ai?
Ig it’s just that my gpu is amd, and it’s not compatible
There is a thing called DirectML where it lets anything that is CUDA to run on AMD and Intel GPUs. Did you download the version that's named DirectML?
AMD/Intel named RVC versions have their DirectML bundled. The speed may depend on your system. Some say it's slower than Nvidia GPU but still usable. 
thanks but i mean the song has to be completely the same except with different lyrics and i know that you have to sing it if you want to use rvc but im looking for something that clones the voice from the original track and everything stays the same except the lyrics
I tried it and ms in games reaches 8k💀
Ayo? @sage vale level 1 !!! 
hmm
Other than the RVC way, I don’t think there’s another way lol
u could try Suno or Udio if they got smt like that, but prolly the best is that u sing it and use RVC
Making an AI parody song to sound very close to its original is quite complicated, since you need a model of their vocals to do it, if you sing out of key you might need to tune it with a program. 
I don't know about Udio, Suno allows you to upload a short snippet of your track for 5-60 seconds and then let it to process. The thing is any copyrighted snippet is not allowed there. 
hi
hi
udio's audio upload feature is behind paywall
No you can with udio or suno but the singer voice will be different so then you use rvc on that
Free trial accounts
bye ijbol this khia rlly wanna fmu 😭 don't make me get oomf to clock you 🧍♀️
Hello
Hii
Wthell the CB 😭
Hey
Hello
Hiii
How are you ?
I'm new here and would like to know how it works
What ?
The RVC ?
Or the okada
The RVC
So we have a complete guide for beginners. Do you want to read it ?
Yes
if u mean the docs, there are also outdated parts btw
it doesnt have kaggles or ilaria rvc zero
- How to use RVC Mainline Colab by Cauthess
- AICoverGen Colab Guide by Eddy (Spanish Helper)
- Create a model with RVC disconnected (colab) by Angetyde
He is new and want to know what the rvc is what he can do etc...
Btw, pretty easily
RVC:
- train (make models)
- Inference (use models) on pre-recorded audios
Wokada: - Inference realtime for calls
Exactly
Ayo? @viral cypress level 1 !!! 
Yea i was just saying tha tthe docs dont have everything
for example it doesnt have wokada nor some cloud options
Well, if he can ask of don't understand anything
Thanks you
Firstly I think he should know what the RVC is. What it needs and what it do
Some people think rvc is good for realtime too, seen people ask how to use applio for realtime 😭
Ahh.. maybe this guide contains everything a beginner needs to know
Btw, if yall meant the technical way on how it works, the docs don’t have that
RVC (Retrieval-Based Voice Conversion) is an advanced AI voice cloning software, developed by the RVC-Project team. It's considered the best free & open-source one to date.
It was designed for desktop, requiring great specs to run it effectively, specially GPU for training models (specifically NVIDIA).
From docs
Im saying that its better to always ask the user what he wants to do and whats the pc gpu
To avoid people using the wrong program with the qrong hardware yk
Imagine you don't know about RVC.. why I'll ask you for GPU 😭😭
if yall mean the usage then yea, if yall meant on how it actually works, there’s a paper instead https://gudgud96.github.io/2024/09/26/annotated-rvc/
because if he wants to know about it, he’s gonna use it
And i seen lots of people with integrated graphics tryna run RVC & Wokada
Bro, rvc is a program use for voice to voice conversion. It requires a high end computer system. But also available on cloud services. What you actually want to do please explain
Im just saying next time its better to always ask what the user wants to do and the gpu the user has
But there is clearly mentiond that it requires gpu specially Nvedia
Is It the right way? @covert lake
You can do it with AMD & Mac too locally, just its not explained here
Yes. I know. Yes I know.
Personally its better to just ask what a user wants and the gpu he has
Anyways dw its all fine
Im just warning you that the docs might not have everything a user want
Its just to not waste time for people that for example would want to use realtime or have an amd gpu
I'm not good at helping... 😫 whenever I try to do that. I make it bulls***
nahh ur all fine
sorry if i sound rude, its just i see LOTS of people using the wrong hardware or software lol
I saw, you ping me many time for this..
I think, we have to write a welcoming guide for individuals that what this server is, why they are here, what they can do, what they requires etc...
technically that’s #1159949278270193734 message
except its all outdated
yo
Who will update it ?
I see. In #1159513888199540817
The domain is expired.. kindly update it... please....
the updated versions of the rvc & wokada collections are https://discord.com/channels/1159260121998827560/1295306886589055009 & https://discord.com/channels/1159260121998827560/1297207135469305866 , but I can’t update #1159949278270193734
So tell me .. what you want to do ?
I’m not the owner of the site unfortunately
It should be temporary, not sure for how much since its been weeks
Just chatting. 🙂
Okay, how are you
If I do /guides
It provides latest link
yea dw ur all good
Well, I'm trying to give my best 🥲 but idk why it always get f#c*ed
No need to stress about it, I was just giving suggestions
Ya ik. You are a good friend of mine

lfg is an abbreviation for that
Idk 🤷♂️ not much active on social media
understandable
@atomic vector how are u
i'm good, thanks for asking how can i help you today
-# AI-generated responses may be inaccurate; please verify important information.
do me a poeam about applio
here's a little poem about applio
applio, a voice so fine
cloning sounds, a digital shrine
it speaks in tongues, a wondrous thing
helping hands, with tech that sings
but for more about this tech so cool
you should check out the main website rule
https://applio.org is the place to be
for all your applio questions, just wait and see
-# AI-generated responses may be inaccurate; please verify important information.
wondering what happened to my training speeds, just upgraded to applio 3.2.6 and i have a training speed of 3 seconds on a 30 minute dataset with batch size 12.

unless you're using like 4090, it seems you may be training on an empty dataset 🙂
I am training on a 4090, before the trainings speeds were 30-40 seconds. not sure what has changed.
I also thought that maybe I made an error with the dataset, but it's all there.
⠀
Google Colabs :colab:
⠀
AICoverGen-WebUI
Useful for making quick covers, by Hina.
AICoverGen-NoWebUI
Useful for making covers, doesn't include a UI, by Ardha, by Eddy, Hina and Gdr.
RVC Disconnected
To train new voice models, by Kit Lemonfoot.
EasyGUI
The OG interface, by Rejects.
⠀
I can't output information to the output screen on mangio rvc, what's the problem?
Ayo? @proud bison level 1 !!! 
delete all .pt files from sileced_audios folder, re-run the training, if a sample gets used in a batch there will be a matching .pt file created
@atomic vector make me a sample text to try a voice model on it
here's a sample text you can try with a voice model:
"hello, how are you today? the sun is shining brightly in the clear blue sky. i'm feeling happy and excited to try out this new voice model. let's see how it sounds"
try reading this text out loud, and then see how the voice model interprets it. you can also try modifying the text to see how the voice model responds to different emotions, tones, and phrases.
-# AI-generated responses may be inaccurate; please verify important information.
Fine. Thanks. What about you?
I'm fine
served CUNT 💜
💀
@sonic slate you got no wrinkles
Bro woke up and chose violence.
absolutely no wrinkles
why are there an uptick of people that just join this server just to spam racial slurs and get banned instantly lol
Ayo? @sonic slate level 1 !!! 
you've literally been in this server for a year and this is what you do when you finally talk
absolutely no wrinkles
If I was a mod I was about to mute you to take some time to think about how to talk respectfully to people.
ow the edge
quite breathtaking
Aceh merdeka
He isn't. 💀
i guess thats it then
suprised the mods are taking long to even get to it but staff inactivity do be like that
fr
Interesting
well now he's banned
AI hub is well moderated
Did a search in the FS discord.
3 deleted accounts say June 12th 2018.
i should have just pinged my god lol
anonymized
i would ping but i wanted to see what he says
Yeah.
In order to anonymise messages sent by deleted users, Discord sets the author of the message to a user whose ID is 456226577798135808 (if you copy the user ID of any deleted user's message, it will always be that ID). Since the ID actually contains a 42-bit timestamp of the millisecond that thing was created, it is also used to determine the account creation and hence join date.
If you plug the ID into a website that extracts the timestamp (such as this one), you'll see that the timestamp is 2018-06-12T22:41:32.427Z. Which is.. close? I imagine your timezone is ahead of UTC by at least two hours, so it went into the next day when displaying it.
This is from a quick google search.
What are y'all doing
What do you think?
I think so.
IP: 125.84.159.131
IPv6: a592:c6aa:21a6:aef8:424a:d88a:2e7a:1ae1
MAC Address: 34:EF:16:01:27:3F
Address (not exact): 1848 Zohme Terrace
Used: /doxx @tepid basin
send irl pic of outside and someone will use their geoguessr skills for evil
rainbolt
IP: 238.84.118.74
IPv6: 9b19:e6b3:dc9b:42ae:46f1:aace:47bc:f88f
MAC Address: 20:35:DE:3F:D0:68
Address (not exact): 6 Dehku Path
Used: /doxx @split whale
Nooooo its accurate
Oh, F. You really found me.
wainbowt
Nah
imma swat somebody tonight
piza
pizza pizza
I imagine the new users joining seeing this and believing it 😭
IP: 120.163.74.113
IPv6: fc5b:814f:63ec:c264:0c88:e687:e674:ff17
MAC Address: C6:39:3B:9C:1A:6D
Address (not exact): 550 Jude Circle
Used: /doxx @covert lake
i mean if it randomly generates their address correctly by chance then they'd believe it
See its accurate
they have not a single wrinkle on their brain if true
I was bored, so I trained Flux some images of Dell OptiPlex PC on Weights.gg. 
I need to check out their image stuff ngl
Looks pretty good
Better UX than sifting through loras on civit.ai
ERROR: Cannot install -r requirements.txt (line 6) and fairseq because these package versions have conflicting dependencies
someone can help me fix this?
Ayo? @daring mirage level 1 !!! 
Why the long cat
😭
What are you using?
And what are you looking to do
What's the Kanye bear?
Been trained these computer pictures on PixAI.art, and the results were disappointing. Better hope it will generate best on Flux. 
rvc gui just install it
Don't, its outdated, NEVER follow yt tuts for this
What's your pc gpu?
And what are you looking for?
gtx 1660 super
just to make some a.i covers
rvc gui is archived lol
i used to work with this but got a new pc and cant download it rn
Your GPU is good enough to do inference (use models) locally (on ur pc), you won't be able to train (make models) but use them
You can:
- Locally (runs on your pc so the speed depends on that, you will have to set it up with the guides):
- Cloud (remote good pc, easier and faster than ur PC but it's limited):
- Ilaria RVC Zero: fastest and simplest that you can get for free
- Weights.gg: Partnered with AI Hub, lets u do them easily but u may be in a queue
- Applio Colab: max 4 hours, not granted, of GPU
thx bro
so apopolio is the simpelest?
Oh god.
Apopolio
That's the crazy name you called.
Ayo? @solar torrent level 16 !!! 
Applio*
If you want to do it locally & easy, yes
the link is a guide
That feels like it's slower than my laptop, damn. 
likely is
Yep.
my laptop has the same specs
😭
i dont know why the fuck anyone decided it was a good idea to put windows 11 on a celeron n4020
fr
with a SPINNING DRIVE
@obtuse wigeon's account was compromised, don't trust any messages from them
Wha
it doesnt have an ssd so if you have windows defender on and it scans your laptop is unusable for like an hour
unbelievable
Bro can NOT run doom /j
Without the /j.
i can literally only use my laptop to play multiplayer minecraft and stream from my pc
GZDoom, you mean. 
hell yeah u can
if i want to play modern minecraft singleplayer, i have to get on windows and play bedrock
People ran it on a pregnancy test
or somehow pirate education edition and play that on linux
im not buying minecraft again just to play bedrock on linux on my laptop which i never use
how tf u pirating education edition
Unfortunately, my laptop has became old enough that Windows makes it heat up to 80 degrees if browsing Discord.
I genuinely wanted to play education edition tbh
💀
i dont know i tried using it once and because i have no fucking wrinkles on my brain i dont know how to get it
so i gave up after like literally 5 minutes of googling
and said this is pointless because all im gonna do is turn on RTX on linux and uninstall after the fun is over
You need like a School Email and shit
That's even worse. 
That's why I use the web version.
yeah
rather would just stay on windows then if i wanted to play bedrock lol
Buy a new one 🙏
Sure thing.
im putting my money on this guy is gonna say a racial slur
Hi there
Hi thanks for the help but I have already done inference with the help of RVC several times. The problem that I am facing is that my computer has become too slow (because of some virus) and it cant even do that. Can you please convert the audio? (I have already trained the model)
You can use cloud services
why not try to clean install or get rid of it? even if my pc is potato i wouldnt want any malware in my pc
Wha
- Creating Datasets for RVC using iZotope RX11, by Cauthess
- Gathering and Isolating Audio, by SCRFilms ❄
- Instrumental and vocal & stems separation & mastering guide, by deton24
- Vocal Mixing Tutorial, by Roomie
- https://mvsep.com/
Oh.
i hit it offline
Ok.
Great
Hi
Hello
goodbye
Does anybody know what programs to use if I want to change the whole style and genre for an ai cover?
Maybe stable audio.
yo
Bro left
hi
The message I sent: do not click on the links, I didn't send it, my account has been hacked.
Hello
Hello
What to do now. I have clicked. It was a YouTube video. I just closed it
wdym?
I got level 30 🥳🥳
hi.
Hello
We use RVC
Okay. That's great. BTW what you want to do ?
Create an Ai cover ?
Ayo? @median cairn level 1 !!! 
Hey guys, does anyone know of an AI that can edit an image based on a prompt I give it? I want to send it a photo of an astronaut helmet I made and i need ideas to make it look broken and bloody, like it was used by a zombie
Where do u guys cover ur ai voices ? I used to do it on Replay
What are your system specs. GPU VRAM
The integrated GPU ?
Like you don't have physical one
no, me poor.
Well which processor are you using
I prefer Applio
Maybe stable diffusion Ai will help you. Or you can visit Adobe website
Intel(R) Celeron(R) CPU B800 @ 1.50GHz
So I recommend you to use Cloud services
is it a web?
No. You will use RVC on cloud computers which are very powerful. You can use Google colab for it.
For free
But have some limitations
like...?
Only 4 hours of GPU Runtime. Unexpected disconnecting
Perfect for making covers
Not for training models. Use can use kaggle to train models
Or use ilaria rvc on hugging face
Quite better and easy
Where can i find it?
no ty
hol on
Ilaria RVC: CLICK HERE 🤗
Guide on how to use it: CLICK HERE 📝
Don't forget to thank Ilaria if you find it useful! 💖
to online 1 live da vida, tema , como ficar rico rápido, sem filtros !
@median cairn
Yw : )
You will also need a voice model
You can fund it in #🔍│find-models or #1175430844685484042
Is this is for promotional purposes?
i alr have 1
Ayo? @median cairn level 2 !!! 
Hes advertising i think
hiii, anybody knows what method of these is better to change a regular speech audio and not a song? (dio, pm, harvest, crepe)
rmvpe is the best to always use
my rvc doesnt have that option. Do i have an outdated version?
Are you following a random yt video?
no idk i downloaded this rvc software long ago
Prob very old
im shocked how it doesnt even have rmvpe
Anyways, what's your PC GPU?
You can locally (on ur pc) and cloud (remote good pc)
You can locally via https://docs.applio.org/getting-started/installation#amd-gpu-support-windows but idk much of amd gpus
Documentation for a high-quality, open-source speech conversion ecosystem designed for simplicity and optimized performance
Or else, just use Ilaria RVC Zero on cloud for using models
Ilaria RVC: CLICK HERE 🤗
Guide on how to use it: CLICK HERE 📝
Don't forget to thank Ilaria if you find it useful! 💖
okie tyvm
RVC is not technically made for that
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.aihub.wtf/tts/gpt-sovits/
Freemium 11labs: A 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
so what is made for? music? i need a good one for audio too, but thanks for the suggestions
It's for Speech To Speech
do u know a good program?
u can technically use it for tts by making a tts audio by another software, then using that audio as an input in rvc
but won't be the best
As i said above, the best TTS are either 11labs or GPT-SoVITS
the notes are different
and the melodies
lol
Hey chat, question
well yea used for instruments too
what is it?
So I installed all the stuff to use the voice changer but I have a GPU, how do I make it work good still
First, what's your pc gpu?
and are you following a random yt tut?
im guessing u tryna do realtime voice changing
A friend helped me
Ayo? @open raft level 1 !!! 
I was trying to use anime voice changers like Itachi but it echoed so badly
what's ur pc gpu
and are u using wokada or the wokada deiteris fork
I'm not sure because I'm not at my PC right now, I'll have to look at it whenever I can
did u follow something like https://rentry.co/ForkVoiceChangerGuide
hola alguien sabe cómo se puede mejorar la calidad de los covers AI, descargué algunos modelos y suenan con baja calidad, en cambio cuando escucho por ejemplo en YouTube los cover suenan limpios y con mejor calidad, ¿cómo se puede lograr aquello?
No, a friend helped me
What
ok
@weary bough Здаров братан какой ставить Tune на твою Russia Women #2
RVC v2
RMVPE
500 Epoch
13000 Steps
вот эту
every time i use a model i just sound like a lil boy lol
Ayo? @balmy basin level 2 !!! 
👋
sup
@paper blaze? 😮
btw What do you think this Gesture is called?
#🏙│ai-images message
Guys, is there a way I can buy an api to get access to newer models for cheap and only for what I use? For example, chatgpt o1 or Claude Sonnet?
Idk if this is right server to ask in
Is anyone on here In the AI lab server too? (I think I got banned there but never got any notifications nor broke a rule and wanna talk to someone on there)
Ayo? @coral ruin level 2 !!! 
Sup, can anyone help me with mlagents?
i don't know how to change cpu to cuda, i have installed cuda 11.8.
I just tried to search for "applio". 💀
@split whale can you code?
@paper blaze Delete that please.
can you code?
?
Why do you ask?
?
Ayo? @lapis karma level 1 !!! 
ok
Stop.
can you?
Ayo? @bold basin level 2 !!! 
Yeah, but I don't know how to code ai, lol.
@wooden jolt can you code?
ok
guys does anyone have really good girl voice model ?
Are you going to ping every member in the server?
.
lol my post got deleted in #1159290752195633273
Now I got to wait 6 hours, I didn't put anything which I would think would trigger an auto delete # -_-
Rather inconvenient
Hey everyone! I’m Marcelo, a consultant and AI enthusiast exploring the latest in tech, especially in AI-driven tools like ChatGPT. I’m currently working on a project to bring AI-powered consulting to businesses in industries like retail and coffee, aiming to make AI more accessible. Recently, I’ve been testing out the new ChatGPT Search feature, and I’m excited to connect with others who are experimenting with it or using AI to transform industries. Looking forward to contributing and learning from everyone here!
yo how can i create song with my own voice, recommended programms? ai sites?
y
happy halloween🎃
any recommendations for best local non-real-time speech to speech ai that supports cpu only even if a lot slower?
Ayo? @ember coral level 2 !!! 
still rvc or applio
u should be able to do it on cpu even if soo slow
hmmm i tried installing applio but it seems i am too silly to understand the process
Go to models and check the vibe
hello all
-rt
This interaction has expired, use the command
/guidesif you wish to see it again.
download the compiled version, unzip, run... how hard is that?
yo i got like an ai question someone with big brain dm me
Hi
HAPPY HALLOWEEN
hi:))
Hi
It’s been an hour people have failed
Ayo? @remote saddle level 1 !!! 
Goodluck soldiers of nnn, this year we may survive
how suggest topic
its 20:11 bruh
1:21 for me man
whats the diff between onnx and regular rvmpe
Onnx for non nvidia
Regular for Nvidia
how do u get it to sound less botty
Use good model and use optimal settings
wdym by optimal
Ayo? @rotund raptor level 1 !!! 
What's ur gpu?
3080
Normal or fork of w-okada
normal okada
use the gpu-cuda installation that has gpu dropdown instead of dml0, dml1 thingy
yes i chose the 3080 as the gpu on it
Currently excited for Orion AI
100X power of gpt 4
But it will probably be even more limited then o1 preview unless o1 preview means orion1 preview
100X power means 100X efficiency
Say o1 preview costs 10X power of gpt 4
Orion will be 10X as good as o1 preview and have 10X usage cap
True but its said it will probably come in December
why do videos use crepe as the f0 setting
Just gotta wait for more than a month then
Because videos are always outdated
ok so use rvmpe
and use forked okada
Yes fcpe and forked w-okada
i must sound like an egirl
Not helping with people that want to use egirl voice models
No you don't
Hi, everyone
Hi.
HI
didnt even do it right
No one here will unless it wad only was a dollar or less so playerup is where to sell accounts
So it's only a dollar the account
q
Ayo? @dense ore level 1 !!! 
I already said a dollar or get out and go to playerup
Can you allow custom reasons?
Anyone here needs help with influencer marketing?
Guys what do you think about the Minecraft movie!!!
i
Hi, does anyone know how to create a cover but as a duo, meaning that two singers sing but in the same song on Android?
Ayo? @honest scaffold level 1 !!! 
Hello, how to get AI voice model changer?
Through #1192011222023950368!
The minecraft movie :O
-rt
This interaction has expired, use the command
/guidesif you wish to see it again.
The minecraft movie looking pretty sick ngl
https://www.reddit.com/r/LocalLLaMA/comments/1ggrwt7/this_is_fully_ai_generated_realtime_gameplay_guys/
Ayo? @spiral timber level 2 !!! 
hi
HAIL H
hai peeps
Hi deer.
im having a blast making retrowave music today. my friend showed me how good the music stuff had gotten and as a musician myself i was excited to try it because i have so many lyrics left unused or ideas for songs i can simply describe to an ai and it will produce now. honestly having a blast making a little ai retrowave/rock ep
Ayo? @marsh flint level 1 !!! 
so far i found that while suno seems to be the overall best it also has some problems such as style consistancy across extensions and also a 130 second memory so if you go too long without repeating a chorus or something i find the ai forgets how its supposed to be sung or just strait changes the singer to a degree if you did a long instramental section.
on the other hand while brev.ai is kind of bare bones and possibly sketchy in their claims the outputs are much longer and more consistant and also it has way way more awareness of what genres exist. other ai singing tools seem to only have the bare basics of genres. most of them had synthwave but none of them could do retrowave ot felt like. and when they did the voice would be way too robotty unless i forced it not to be.
overall i think if you know what lyrics you want and what genre then brev does the best singing and very good instramentals(albeit with less variety than suno)
@bold rampart its fine to put the covers there id assume but they may not embed and even if they do only 1 post every 6 hours in that channel.
Ok
hi
Hlo
does anyone know how to make isolated vocals cleaner for ai? im trying to make a fake finalized version of a song using filtered vocals of the writer's demo and a model trained on a snippet of the final version... problem is the vocal mix of the writer's demo is a bit weird and makes the ai output sound muffled and full of artifacts...
Wdym weird? Is there noise or anything?
basically the eq is like all messed up
Hello im new
Hi, welcome to Ai hub
Hi new.
The one that says hi gets: ||nothing.||
Ayo? @elder willow level 1 !!! 
can you send me audio file in orginal + ur isolated?
in dm
ha
ha
yahh
hello
Hi.
Ayo? @split whale level 9 !!! 
bye
says some error about gradio not being found


AI HUB Docs