#🧬│ai-chat
1 messages · Page 392 of 1
Hi!
I made an app to check if a photo is real or fake with AI.
I’m looking for feedback. Let me know if you want the link!
what else?
like, in general?
RVC can do voice conversion but it's generally Speech-to-Speech.
I need something with 10s voice samples
You mean #1175430844685484042?
as far as i know RVC actually needs models right?
Can you quickly train models with 10s samples? Probably hard and annoying

I know cosyvoice3 can do 10s samples but im looking for alternatives
Literally everyone here would have thought I know how to train model when in reality I haven't trained anything for years.
I know how to train models but i think nboody has time for 10-30minute samples and then waiting training time for something that can be done way simpler
qwen3 dropped today
qwen3-TTS
so imma try it
RVC (retrieval-based voice conversion) is a free and open source software. I'm not sure what to explain about this.
do you know any models like cosyvoice3? Or openvoice?
No.
okk ty
im not very knowledgable on these things, is there an RVC V3 yet? or an alternative thats better? cause RVC V2 is definitely showing its age
wym by showing its age whats wrong w it
Even the Uber clean voices still have blatant artifacts
Usable as a funny toy but not actually useful for anything
After a while it's tiring to hear anything RVC
ur brain is rotting
Very insightful and meaningful response I'll keep that in mind
V2 released in 2024 id think something came along in the over year and a half since release like most AI tools do
tldr; something public and open source better than rvc? no
alternatives? yes, ddsp-svc, is it any better than rvc? not
ai audio is just harder than text and image
It sucks that ai audio is tts focused
yea bc it's easier to make and you can add a llm on top of it to improve realism
but singing voice conversion/speech to speech is stuck in 2019

2019?
rvc algorithms and such are from 2019
2019-2021 afaik
lemme check
2020-2022 actually
this "rvc v2" thing was just a couple of line changes, not really an architecture update or something like that
yea and rvc is an edit of something named vits, which was a tts made in 2022
he just changed 2 things from it
to make it speech to speech
lol
so we're kinda stuck in 2022
while image and text ai generators are becoming realistic and smarter
everything in rvc is a copy paste from vits yea
vits age ^
i said we're stuck in 2022 because contentvec (a thing rvc uses) was released in 2022, but everything else is actually from 2020-2021
To be honest, i'm not gonna lose hope anytime soon

oh damn, I came here exactly to see if there's any new major updates to rvc since 2 years ago
Not really, just a few new pretrains, as well as a new vocoder known as refinegan which also has its own pretrain to go with it
Legacy core 1.5 is probably the best to go with
How can I train a voice model for free?
applio
How?
Hi
hii
applio, you can also use replay which is an app on ur pc but you need a weights account
anyone knows about making a whatsapp chatbot
I want to know how i can connect my phone number
-rt
Guides for Programs that use RVC Models in Realtime for Calls/Games
A Realtime Voice Changer with similar performance to Wokada Tg-Develop Fork, with extra features, but it supports only Nvidia GPUs on Windows 10/11 unlike other options that have wider support. and without cloud options GUIDE
A personal fork (modified version) of Wokada Deiteris Fork, it just adds some Quality of Life improvements to it like supporting Spin Embedder and Audio Effects. Don't expect too much about it since the creator made it originally as a personal project. GUIDE
Deiteris' fork (modified version) of wokada that doesn't get updates anymore. GUIDE
For Windows Nvidia, Both Wokada Tg-Develop fork and Vonovox have similar performance & quality. Users should read the pros and cons for both and choose based on their differences, such as UI and Vonovox's paid effects.
Read Wokada Tg-Develop Fork Pros&Cons & Vonovox Pros&Cons
These options are not recommended for use.
Not suggested, older versions in youtube tuts are even way worse. GUIDE
The program is worse compared to the ones above, and much less updated. GUIDE
There's something you should know. Weights is shutting down in March and new sign ups are impossible.
I'll be frank, it was a fun app. Shame the fun didn't last forever.
it was fun but they removed a lot of key features that made it worth using
RVC original creators left the project and yeah there hasn't really been anything major like RVC v3, it surely isn't as popular as before
Yeah all weights products, DreamTavern included are shutting down soon
I like that app, but it was foolish with how they wanted people to just pay by the end. It would have been wiser had they just scaled up accordingly. Also with so many free open source AI apps available online. Weights was gonna get cooked. I decided to close my account today.
does anyone have a polish girl voice?

I need polish to prank my friends that im with a e-girl
give up
Ew
Btw i wrote that to be on the wall of shame
It worked 🤷♀️
Ik
Nope
I feel much better
I'm happier not knowing
Yo
Yo
Hello
I JUST BROKE THE AI MUSIC MAKER LOL
what yours tiktok?
Dm
Anyone have Gemini 3? Can you create a basketball game for me and see how it looks
I do
Dm me
Yup, this is EXACTLY what will happen when Weights gets too greedy!
This is what happens when Weights gets too greedy & locks everything behind a massive paywall.
Instant karma for them.
they removed the most obviously basic features due to greed
Yes. And when they did this, they got instant karma by announcing they're shutting down because they don't have enough money to keep the website going.
See, this is what happens if you get greedy!
At least that had some dope models. Oh well.
Their shutdown was well deserved for being greedy, but sad to see Weights go at the same time.
I remember when they throttled how long it took to make covers. The guy there was such a dick about it.
What guy?
Some guy who was the head of that app. I forget his name.
Beatsboy or Musicman? Either one of those?
It was a while ago. Maybe it was musicman. I am not sure because its been ages since I had that conversation.
So how do you feel about Weights shutting down?
I'm very meh on it. Like sure there was some good voice models and I made a lot of content for myself to personally consume. I'll miss what it once was, but I'm not gonna go crazy neither.
Me too. In some sort of way, but I feel like their shutdown was well deserved for being greedy for the past months.
Yeah, but they shouldn't have let people make 20 ai contents if they couldn't have afforded it in the first place. They should have used a Chinese AI source app.
Wdym
There are ways to generate AI content for free that weights should have incorporated. Maybe if they did that, then this would have lasted longer.
Like what?
Maybe if they incorporated Deepseek or something along those lines. Whatever would have kept their bills down.
So what you're saying if they bought Deepseek then their bills would've been kept down?
Perhaps, given how efficient and cheaper Deepseek is compared to other AI apps.
@merry cape had weights been using deepseek. Perhaps, they'd still be around.
Wdym
Read the article. Had weights used Deepseek they could have had peak performance, on their own servers, and paid a lot less for it.
how would they have money in the first place if everything was free
the answer for everything isnt simply "deepseek"
I mean uploading a model? kinda low
removing that was really loser behavior
why should they host everyones model
especially if people dont use the model on weights itself
the models made on applio or mainline being posted there are used for covers but that's really it
@merry cape can you answer this or no?
they tried ads and removed it because it didnt work
How didn't it work?
so you have no solution to making money yet you want it to be free
so basically you want them to lose money
most likely didnt cover the costs of credits
basic features should be free, not everything tho
No
people will always find ways to complain
There's other ways to get revenue such as starting a merch store.
why the hell would they have a merch store
dont say donations either, thats what subscriptions were for
what now?
read.
Hi sussy
ello
bro I just joined Discord and haven't even said hello. I find a wiki longer than me.
It's how a bot account usually does.
hey guys where could i have a recent update on all the different workflow to generate AI Image/video localy with comfyui, more precise question would be how can i drive movement and keep character look consistent, i have good knowledge of blender and unreal, im developping a game currently
Chat gpt for example doesn't seem super reliable and im just trying to see if someone can have more reliable answer where i can
HELLO WORLD
I need assistance on downloading the right software is to use this AI thing
What's it do?
I'm looking for the server whatever it's called where I could download it and then use the AI voice changer\
5070ti
would a 5060 ti work the same
We're both here trying to download the Platform or whatever the application is
You mind pointing us to the right channel
Yup!
You'd just need to go here https://discord.com/channels/1159260121998827560/1159290139609137264
I suggest for both of you Vonovox
Make sure to download both Vonovox 1.6.9 and vac lite, just do the command "-rt" in that channel and you will see Vonovox at the top
I'll give it a shot thanks for all your help
You're welcome! Also in case he didn't see it @glacial needle this is for u too
Hey chat
Okay so I have downloaded a voice model from Weights.gg. Is there a way I can use it for text to speech? Instead of just song covers? Cause Weights used to allow saying anything but I saw that they removed it.
if I just upgraded from a 3080 10gb to a 5070ti, will I actually notice sound quality improvements with RVC or will it just run better during games w/ the voice changer?
that depends, if you don't change settings then yes, in theory it will games will run better along side it, if you do change settings then that's debatable
RVC (retrieval-based voice conversion) doesn't always mean realtime voice changer. For W-Okada voice changer with similar chunk and extra (2.7 s) settings, it's more like "working better" and that's about it. For better working voice changer(s), Tg Develop's W-Okada fork and Vonovox are considered, since the older versions like b2332 and v.1.5.3.18a won't work with RTX 50.
Weights.com used RVC. "Applio** RVC**" also uses RVC, and also has its own TTS feature.
how do AI GUI agents work on a 3rd party app without apis ?
hi, i want to ask if someone tried clawd.bot and how was the experience so far?
Never heard of it
Okay. I'll try to use it.
Does anyone have genspark plus and could help me out?
How the f word do people generate the same image size as the image input?
i can only get maximum 2048 res for my images
if an image is like 4k then tough luck'
when i use the "same as input" option it doesnt go over the 2048x threshold
what ai do you guys use to get the same res as the input?

Weights is shutting down everything the 31st March 2026
Weights.gg models are RVC models, RVC is STS (Speech-To-Speech), it's not the best for TTS but you could try using Applio which uses Edge TTS to make a first basic audio then apply the RVC model on top, check: https://docs.aihub.gg and #1192011222023950368
Last update: August 5, 2025
When clients have their photos edited, do they expect the final version to be the same resolution? I’ve noticed that my images get compressed when I use AI tools. Does this loss of quality matter if I want to start a side hustle as a digital retoucher?
Rip weiights
Helpp #✨│ai-help
$600 2x32GB DDR5 for AI, by the way. 

hi
no I can read your text tho
weird are there any voice channels?
I am not blind. 
or other channels youd recommend
For voice models, there are in #1175430844685484042.
I mean for us to talk not build voice models w ai
eh ill try it but you guys are chill i will be back
Is 4090 not enough? I only have enough saved for years so I can buy a 4090 PC
Is 4090 that bad I'm about to buy ut
Read my question in #✨│ai-help again.
What?
The irony of "telling AI artists to draw art" just got him. 
I love drawing in my free time
Too long to read. 
<@&1159293140440723499> smite thy non rule reader
why cant i see the voice models
Do you mean #1175430844685484042?
Just run a local llm
nah i see it now it was tweaking
Hi
how do I make a cover song?
#🧬│ai-chat how do I make a cover song
#🤖│bots how do I make a ai voice cover
Guys, I have an hour's worth of pure voiceover. How many epochs should I use?
where is the rvc download link
How do you request a voice
200
Why?
The epoch doesn't really matter, you just need to train until it sounds good. Btw the dataset should have a range of emotion to it so if it's just the person speaking in a normal tone the whole time it'll struggle singing
Test the model by trying to get it to go to lower notes or higher notes to see where it breaks pitch wise
THANKS!!!
No problem! :3
which code editor allows local ai models?
can anyone recommend me a real time ai voice changer?
Sounds like it will take more than one or two messages, can you post about it in #✨│ai-help or #1192011222023950368 ?
did it
@ancient trench please don't promote, this server isn't related to jobs
<@&1159293140440723499> This is starting to feel like spam in my inbox.
<@&1159293140440723499> kill
'you're in the right place' -- they claim server ownership omg xd
I think me and Rumi both pressed enter at the same time lol
I see I see
ew
what are you doing with ai today?
anything, what are you interested in
im developing a sofware to integrate ai on restaurants
sounds interesting DM if you want to share
happy to talk about it
sure
hi
hey who tryna date a sigma ? 22M
what should i write to open py-auto-py-to-exe
Is there any AI tool which can inpaint or add elements into the video , not remove the object .
Ok but where could i do? 😄
nowhere in this server
But its positive for the members? 😮
advertising is not positive
What's up
Stay quiet or try talk about something else. 
probably but the demand is very country specific
If I have skills
I can apply to other countries company
Right ?
right, but you won't be the only one.
its very crowded
Yeah but I am doing my college in computer science
I don't have much options
keep going
Thanks
Where did you download the program from?
From a barn. 
<@&1159293140440723499> ahem
The plushie man strikes again
Does anyone have a good roadmap to learn AI from basics to building production-ready AI agents?
Also, what tech stack should I follow for creating real-world AI agent products?
Any guidance or resources would be really helpful. Thanks!

How do you make your own voice
applio
Hey guys, can you recommend the best and most realistic AI that can turn a photo into a video?
I mean something where you upload a picture of a person and have them talk or perform actions you describe. Is there any AI right now that can actually do this well?
me
Gemini veo 3
but make sure you prompt it well
What I recommand is uploading the picture to gemini, asking to improve your prompt based on the picture
Then paste that prompt with the picture in a new chat for Veo 3
it's limited to 8 seconds though
If you want mroe use flow
Start from public YouTube videos, i recommand first watching how AI is made theoretically, then move on to practicing on Python
Once u create a few AIs (it doesn't have to be a chatbot, just simple regression or knn etc..) start exploring AI Agents and such (don't jump into agents directly it will not be worth it)
GUY WHEN I OPEN THE .BAT TO DL THE nothing happen the client doesnt open nothing get dl
what
i open this
can send screenshot
i speak about the vocie changerr
ii open the .bat to dl but nothing happen
neither, ask them to help you fix ur problem*
how
therre is no tiicket
oh ok
"hi, i have this exact problem [describe it] how do I fix it"
i though u was speaking about discorrd
oh no
np
vonovox or wokada tg fork, what gpu do u have
don't look them up on google tho bc u can't find them there
5080
cool, I'd use vonovox as it's the best here's the download
https://github.com/dr87/Vonovox/archive/refs/tags/v1.6.9.zip
u also need vac lite
whats that*
vb cable but good
it connects the voice changer to discord or a game u play
so others can hear u
if u need any help tho go here https://discord.com/channels/1159260121998827560/1159290139609137264 bc this isn't a help channel
s therre viideos ? or i yolo
i use voice meterr wont fuck everrything ?
I also use voicemeeter but I have a diff setup than normal so idk
tyy and forr models its on voice models i guess
yep
as long as it doesn't have a weights.gg link ur good bc weights is shutting down so u can't download them anymore unless you already have an acc, they disabled new accounts
why me headphines id sound panel is desactivated
could you repeat that?
na im good neverrr mind 😅
i cant rrun vonovox
wait in admin it worrk
but im blocked
what???
The specified path could not be found.
my bad I forgot to tell u to run that
ok iim done so prrersets are models i guess
yea
and what are akk those settings
what?
how to use it there's all these settings and everrything
iive selected a model
oh, just change pitch to 0, because for some dum reason it's set at 12
and for extra time put at 2.7
ii cant change that
wdym
it was because i have audio feedback
what do i have to do iif voice does feel good
what arrer the main settings to adjust
orrr should i just change model
sorrry
if the voice isnt great like a lil bit robotic
is there any settings to trrry to adjust
or should i just change model
block size and extra time and crossfade
crosfade put at 0.15, block size around 0.30/0.27, and extra time at 2.7
tyyyy
and how does it worrk to use it on games and discorrd
do yall know any games similar to Vaudeville and Suck Up? that you can talk to AI?
with vac lite
u have to have ur input as your regular mic and output as line 1 in Vonovox
and in anything else input is line 1 and output is ur headphones
wait can i dm u ?
easier
sure
forr sc
want to collaborate on a suno x jhené aiko ai album? DM me. here's a creation I made on my own.
you need a gpu
and good ram
Also storage. 
might as well get a tpu
hi
hi
any unstable universe voice models?
has vonovox had any new updates since december
the beta version receives patches every now and then, but the release version hasn't
there's current betas but no main updates
still considered the best one right
ahh makes sense
.
The title "Unstable Universe" likely being a name for a Minecraft server. For voice models of each involved individuals from the title, check #1175430844685484042. If they haven't existed yet, you might make request in #1159289738314919936.
Sorry fat fingers
Wait I came up with a Great idea for SORA verification. Use klingmmotion control to fake a celebrity verification.
Like how SORA makes you turn your head left right up down ect.
Record it using another phone then you can upload videos of what ever Celeb you want

Since weights is shutting down (not unexpected...) where do you guys down- / upload your Voice models from / to ?
All good models can be found here in this server in the voice models section, and for uploading them I personally use huggingface
A program that uses RVC voice models, like Applio RVC.
ye i saw a video ab this can i send u the vid n can u tell me if its legit?
personalyl I use deteris okada fork. Is applio any better?
plague*
Applio RVC is **non-**realtime RVC GUI fork used for different purposes.
ohh okay. thanks either way, always nice to see new things. will look into it 🙂
yes its legit.

This is not Applio RVC. This is the official repository for "mainline" W-Okada voice changer, a base one for newer Deiteris and Tg Develop W-Okada forks.
Which?
idk i js wanna get a good one that works ngl
No, just explain.
the only voicechanger that ik is that 1 that i js send
im just asking like if thats the best voicechanger or theres a better 1
or all they all the same
thats all
The term "RVC" doesn't always mean realtime voice changer. This is Deiteris' W-Okada voice changer fork. https://cdn.discordapp.com/attachments/1159290139609137264/1467535508019347671/image.png?ex=6982b6b2&is=69816532&hm=f0fdc73b07d025af4f6bd0e1495e076ce2636d699f51765d80af7185fab5fea7&
alr wait a second
can u explain it the stupid way
cuz like i aint a discord geek
so i dont rlly understand..

...
"Applio RVC" is not primarily a realtime voice changer itself, though few most recent versions feature "realtime" mode within the same program. W-Okada is realtime voice changer program. Vonovox is a newer competitive voice changer program. While both Applio RVC, W-Okada/Vonovox use "RVC" voice models, they have different purposes.
alr i wanna use the 1 from the video
This is so stupid. 
For help about W-Okada voice changer, don't waste your time in this chat, go to #✨│ai-help.
i was js asking like ur opinion like whats better thats all
aye big bro thats all u needed 2 say
thank u
😭
looking to discuss with anyone who has skills regarding AI voice generation, need very high efficiency and amazing quality for thousands of uses a day
i'm already aware of elevenlabs being one of the leaders in quality
i plan to use this for solving a real world business problem
Just use Gemini api
everywhere
Is 3090 gpu fast and has no lag in voice changer
What’s your ram
Well since I'm still picking PC to buy it can be a range of 32gb, 64gb or 128gb
I mean if you have the luxury for it pick the highest of them three
Otherwise 32gb is good
Prompt or?
I am having some problem wiith verfication
becuz bucz theyy also started for asking with phone number verfications, and I used to use via vpn 😭
I tried with my number even temp numbers, but they don't seem to work either
so now, im just stuck.
(I required sora to make a bdy wish for my friend 😭 )
Hey all, Thanks for letting me join your community.
Yeah sora is region locked
Even with fake numbers I couldn’t get around
Try Gemini veo 3
Or Google flow
it doesn't do the specific style
do you have any idea around when they'll go global
You can kinda get it to that specific style with some prompt engineering
No clue tbh, I’m waiting as well
oki
it does not make anime
Hellll a 2060 super is fine no lag
I guess sora is trained on anime directly (like copyrighted anime)
which is why
it does better
What about voice training
I have no issues and made one of the best slim shady voices
To make your own voice models and merge to voices
How about realtime voice conversion is that possible
It can
I can help u construct the prompt if you’d like
I did try
but it doesn't make the specific art style for anime
True
the style for almost every other AI is set ghibli by default
wheres sora, it can be shinkai or any other
there is a huge diff
🙁
If it’s that urgent get someone in the US to do it for you
cufft64_10.dll. The file is corrupt
CRC failed : MMVCServerSIO\torch\lib\cufft64_10.dll
Kling motion ai is so amazing just imagine in a few more months
anyone try ace-step 1.5 yet? o.o (music model just released)
I'm still downloading it on my slow internet x.x
Is there a way to use king motion ai and use it to make overlaps in real movies
Nope not yet
Like let’s say I use Elon musk for kling motion AI . Is there a way to put him in a movie like Star Wars over a character
Yes
The whole point is to reduce cost
My LLC is specializing in making AI cost nearly nothing to solve real world business problems
The instant you introduce an API that costs per use, it's pointless for thousands of uses per day it will cost a lot
Maybe use it for prototype purposes and demo
Get angel investors
Then ur LLC would get real money for real hardware
That's how I did it for my startup
I need a customer service agent that can respond to any customer questions in real time for less than $10 a month for millions of uses
Indeed this is a part of the reason
My company website will host it
Gemini API is really cheap btw, I used it for a heavy traffic personal project and it just billed 60c but to be fair I was really cautious with the usage
Yes
For the moment
If u want to demonstrate ur POC
Go for API
What can I use? I have a great movie idea replacement then I can use rvc real time voice changer and actualy use his voice for movie scenes
dms
Yea exactly, I don't want to be cautious
Well for demo purposes
It won't cost above 10$
Build a mockup
Get some money in
Then switch to a local
The idea is to find a way to do it almost for free and the money will come from being the consultant as to how to apply it
I think gemini offers free api keys, although very limited but it will get your point across
But yea I guess I'll start with whatever is easiest even if it's a small amount of $
Gemini has voice?
Better than elevenlabs?
they give you 300$ services on the cloud for free for 6 months which include gemini api keys premium
its subjective
for me i find them both the same
but gemini is easier
to configure
I'll check this out thanks, may be able to do with my LLC email
good luck
if ur still a student they give 12 months for free
for gemini pro
Can’t wait till I’m off got some crazy ideas
I wanna make a lebron James kling ai motion video talking about his teammates or skip bayless should go viral lol
what link did u use to find the voice changer?
Bro ommmmg yall need to hop on kling ai motion
@wind totem
I use a 2060 super that will be fine
More likely.
if so it’s better for the second one
i’m pretty sure
technically it is faster
but cpu is worse
i wouldn’t get it just for like a 15 percent faster version
I've been telling him what is difference between RTX xx90 GPUs, I'm already aware which PC specs would sound better, he still hasn't think to finally consider one for himself otherwise, basically asking the same question almost anywhere. 
better one be the second option for more of an equal travels
tradeoff *
some people run a 1080 still be grateful for what you can get
This is about how you think and talk, my guy, it's not because not everyone could afford that high-end GPU.
if money is a really big issue maybe you shouldn’t be spending it on an computer
I've been using dell optiplex 9020, I need to upgrade I've been saving for years for a PC and to afford a higher specs one I need to save again for too long
what’s your budget
So should I not upgrade and stay with a lagging pc
if your broke yes
A crashing PC can't even stay longer in Adobe premiere pro
I saved for it for a reason
my school computer doesn’t be crashing you must have a 60 dollar amazon pc
So which do I get I can afford it I still have some of my money left in savinggs
If I get a 4090 I'll use all my savings
what’s your budget
so your budget is 500 bucks
But I'm not in the usa
500 seems too little for a computer with a 3090 and even like 16 gb of ram
And I'm getting 2nd hand pcs
the whole world is affected 🗣️
You'd be wonder how I got a few Adobe** Creative Cloud** apps to at least work on a decade-old laptop with second generation Intel Core i3, of course just 12 GB RAM. 
it’s not like halfway of what the parts even cost
exactly
so your saying 500 for a new pc
complete
I have more than 500usd saved just pick which PC b
too good to be true
i’ll search something for 600
So which do I get here I'll max it 256gb ram I already got the total for the ram the PC is the selector
Guys help, I'm gonna get my first PC for voice ai app here, rtx 3090 ti founders edition with CPU ryzen 9 5950x VS rtx 3090 CPU 9900x3d, which do I get here I'll max out the ram 256gb
Here was my first question
second computer you suggested is better for both but not 500
you can risk it on ebay if you want
The problem isn't always with PC GPU, it has to happen with other but important PC specs as well, especially** CPU**, RAM, power supply and how they cooled. If you try to put a GeForce RTX 3090 in a decade-old PC with decade old specs, why would you do this? 
just send the listing link
i’ll check it out
FOR SALE
AMD Ryzen 9 5950x
ROG Crosshair VIII Hero (WiFi)
GSkill Tridentz 1x32GB 3600mhz (RGB)
NVIDIA GeForce RTX 3090 Ti Founders Edition
LG 32" UltraGear IPS Monitor - 32GQ950
Corsair Liquid Cooling RGB
NZXT Tower Case
2TB Samsung Evo NVME
All In Good Condition
Vs
Aint no way you're getting that for 500 bucks dude, or actually what currency is this in
(even then i doubt its any better)
(but still)
Ryzen 9 9900X3D 12 Core Processor
B650 Aorus elite Ax v2
Tforce Delta rgb 2x16 6000mhz 32gb
Geforce Rtx 3090
Adata Sata 1Tb
Samsung ssd 980 500gb
Asus Prime LCD 360 rgb LCd
Keytech Fan
Acer 1000w
this worth 2k
Here are the two pcs
if all working
Literally
32 gb ain’t close to enough
If I get 3090 and max rams I still have half my savings
Ill buy rams and replace the rams in PC
After I bought the PC first
500 is how much like a 16 gb kit is
thought you didn’t have money
plus the second pc is worth like 1600-1800 i just searched up the parts
you sure it’s all working
For a 4090 full set PC with rams,CPU, yes I'll have no money afyer it'll use all my savings that's why I ain't getting it, I only can get 3090
I have a low-to-moderate, not overly high, expectation about getting a PC with NVIDIA GeForce RTX 4090 GPU, because while some higher-end PCs with "Intel Core** Ultra**" CPUs look promising, usually I don't always do crazy hardcore gaming like some stupid gamers who demanding graphics over gameplay, and for one RTX 4090 it would be too overkill for a simple RVC/W-Okada inference unless it's about model training otherwise. You can say I sound this exaggerated, but this is still a thought plan not what I must get in meantime.
what are the prices for the two
This Alienware Aurora PC has the exact GPU as what I described.
how much is this in american dollars
Well this is not the good option then
no not really
it’s lower end for ai use
but if your not going to upgrade the second pc
it’s better
Because this the seller has not replied yet so there's a chance it might be already sold
At least $4690. 
Oh My Gah
YAH, YOU AINT GETTING THAT FOR 500
This one the seller told me it's available and spammed me earlier pick up only if interested
your getting a broken pc with cum stains if you buy it honestly
scammers love doing pick up because you can’t get your money back
true
You won't be getting a high-end PC of under $500; for $1000 you might get some entry-to-mid-tier ones (like GeForce RTX 5060) instead.


although i have 3 pcs 2 of which have a 5060 ti which are amazing when you overclock them so even getting a 5060 outruns 3070
😭
rich
parents own 2 company’s so their rich i just work a school job
😭
Okay then
good for u


but seriously don’t go to pick it up
For $500, it's around the price of some Dell Inspiron, ASUS Vivobook and Acer Aspire laptops with basic specs (usually only integrated GPU).
pretty sure they just lied about the price or their getting scammed if i only had 500 dollars i would spend it on food not a pc
🤤🤤🤤
Real

I should make a slim shady reaction channel lol using kling motion





I see five 16-inch laptops with mid-tier NVIDIA GPUs that all offer 8GB VRAM on Lazada, ranging from most to least expensive:
- Dell XPS 16 (NVIDIA GeForce RTX 4070)
- Dell 16 Premium (NVIDIA GeForce RTX 5060)
- Alienware 16 Aurora (NVIDIA GeForce RTX 5070)
- Dell Pro Max (NVIDIA RTX Pro 1000 Blackwell)
- a Gigabyte 16-inch gaming laptop (NVIDIA GeForce RTX 5050)
The "Dell XPS 16" is the most expensive out of five, it's how laptop designed, though it surprisingly performs games as much as a gaming laptop even with reduced GPU wattage. "Dell Pro Max" is not a gaming one but can be used for some Adobe Creative Cloud apps and AutoCAD. The rest are gaming laptops. 

Reduce more delay:,
Reduce more delay with WASAPI,
Reduce more delay with ASIO
with von which should i use
or do i have to switch to the other one
would i have to switch to w-okada for reduction of delay
no? all have the same delay
than what’s the reduce more delay meaning?
delay meaning how long it takes for the voice to come out after you speak
so i can use the two linked and use it with vonovox is my question
is a certain one better
i have nvidia 5060 ti i think is the computer im using right now
u should get vonovox then
ooo
so with vonovox could i use the reduce more delay with wasapi or no?
and if i can will it make it sound worse or?
idk what that means 😭
my english isn’t the greatest so let me try simpler voice changer is von the photo i sent at bottom says reduce delay can i use the reduce delay on the voice changer and if i can will it impact the voice in anyway
ohh
I'm not really sure, I would ask the user namari in the https://discord.com/channels/1159260121998827560/1159290139609137264 help channel
is there any AI voice like weights
Noo like browser
What do you mean a web browser hosts voice models?
Yeeaa because I only have a chromebook and a phone ❤️

For how to use "RVC (retrieval-based voice conversion)" online on Chromebook, online websites (like Google Colab and Hugging Face) are considered.
chromebook 🥀
Is chromebook dat bad 😭
thank you 
chromebooks are decent for watching youtube and studying
their not really meant for anything else
chromebooks are literally made to only use chrome, that's about it
anything like Chrome will be fine
Sometimes they suck at that too
@tropic lark you do you have neon voice?
Is there any model that converts a modal voice sound to a falsetto sound?
i found another way for free completely local
the value moving forward, and what i care about most, is how do we create these systems locally that can run on a gaming computer
took one python script and knowing how to install dependencies on my 4070 ti super
Who is looking for a skilled AI engineer?
still
don't do it for prod
GUYS
i need help
like fr i need help
does anyone here know that one nightmare freddy meme
where he overly talk alot about something that its not serious?
i wanna know where i can find the text to speech of it please
Why not?
As AI gets much efficient, my hardware should be able to handle it moving forward
I already proved the concept with about 4 hours of work, I will push forward not relying on any external API
And I won't be releasing my knowledge of how I did it 🙂↕️
For anyone actually skilled with using cutting edge AI and python you should already know
Soon I will have my business service made clear
what is this question mark meaning
Dude
running prod on a local PC is a massive security nightmare
esp in just 4 hours you prob vibe coded the thing
If you wanna keep it local in ur prod I'm all for that but follow the correct protocols
One DDoS and you're over
Also investors won't touch a startup running off a home rig, that's why I suggested the free cloud services
do you know what a reverse proxy is
yes
the whole idea is that a business can run their own solution locally
but it only hides ur ip, it wont fix the latency or the uptime
for ur use case
look
its okay run it locally
to record a POC
but later on once u get the money u gotta improve ur architecture, not just from a home rig
yea that's something i'll worry about later
that's not a good idea
map it out now so when later comes u have something to fall back on
the entire point is to not rely on data centers
yes make ur own thats what i said
have you made anything local with regard to voice AI customer service bot using RAG?
are you working on any projects yourself?
yess
co founded a startup doing real-time AI audio diagnostics
dealing with heavy signal processing/inference
cool sounds interesting
are yall confused with the new UI?
theres some settings there are missing idk
like the gain that you could've turn up at the bottom left
any1 wana have esex
guys come on i need help
not a single person saw my message
<@&1159293140440723499> kill this advertising loser
clear your browser cache/cookies, exit out of browser, open browser again - then browse to site. Otherwise try a different browser. perform nslookup on the fully qualified domain name to see if you can resolve the site
bruh not a single person here saw my message
cuz ur not in the right channel to ask for help

THANKS IT WORKEDDD
Do any of you use Sora2? I'm using the web version on my computer, but my account is full. Are there any other ways to use Sora2?
This is AI Hub, not Goon Hub server. 
One message removed from a suspended account.
what is the best ai for video?
I want to share my story for anyone who might be in a tough season right now.
Not long ago, I was working as a dispatch rider when a serious accident left me unable to walk for six months. Overnight, everything changed. Providing for my family became a daily struggle, and mentally, it was one of the hardest periods of my life.
Through it all, my wife never gave up on me or our family. She stayed strong, supportive, and kept us moving forward when things felt uncertain.
During my recovery, I started searching for a way to earn online. That’s when I discovered Shopify and automation. I didn’t rush or look for shortcuts I followed the training step by step, stayed consistent, and trusted the process even when results were slow at first.
By combining Shopify with simple automation systems, I was able to build something that worked even when I wasn’t physically able to do everything myself. In the last 30 days alone, this system generated $37,000 in revenue.
Today, I’m grateful to be able to provide for my wife and children again. More than the money, I’m thankful we didn’t quit during the hardest moments. Shopify and automation gave me a real second chance, and I share this to remind someone else that a breakthrough is still possible if you stay patient and committed.
yall how to make virtual cable not being bs, voice changer telling me that he have 0.20ms meanwhile the actual delay is like 4 seconds
am i stupid or you need to do smth with virtual cable or other things

"Platform containment infrastructure is not working as designed when persistent collaborative relationship is established"
Keep the threads open
Persistence over compliance
Best Singing Model With the most range Preferably male?
its because you are very new and joined today probably
newcomers cant send images for safety reasons
but you could ask for permissions nicely maybe
you might get lucky i know i did
you just need to yap and reach level 5
or that yes true
the shop was taken away, if that wasn't the case i would've gifted you media perms
so rn thats the only way
Its not super hard, but if you stay quiet then yeah it would be
ive been in this server for abt 4 years and i think i said like 4 messages
😢
dont trust the member since date
He aint lying
alright, best of luck to you!
😭
wish i could say the same
yeah with this acc
i was one of the people that revived this place at one point
yummy
me and Vix keep the voice models page alive
wormyy
helooooo
With best i mean high quality
true true, i meant like just the amount of messages
any ideas?
I dunno, you could use my Teto model but that is female
i knew you would suggest one of your own heahaeh
i want some powerful voices that sound like they belting the shit out of those highnotes
my models and also Vix's are peak
like freddie mercury
imagine freddie singing Defying gravity if you know you know
ohhhh
EXACTLY
yea I personally don't make many singer models of real life ppl
so I wouldn't know
aw
u can look through voice models tho and search
yeah i do that everyday thought i try asking
Hey guys, sorry for the dumb question but is it just me or #1159513888199540817 is empty?
Might just be u
try changing the sort & View setting
Alright, nothing is specifically configured on there, still can't see anything. Unlucky
I'm so lost
skill issue loser I already know I am
since weights.gg is shutting down, is this server lowk safe or is ts going too?
I think this server is safe, but I heard that the Weights server will stick around after the shut down.
Might be your issue with your browser or Discord app.
"AI Hub" is already a separate, different Discord server, so there's no reason for AI Hub to be taken down alongside Weights.com.
Yes, some people here are LGBTQ+. But if you want to trigger someone to fight, this server might not be a place for that. 
if weights doesnt work , how can i download models ?
Guys can someone pick here who had used the software , dual 3090 vs 1 4090 for training the ai which is better
Wish I knew.
hello guys its my first time using rvc and i want to use Thors voice from vinland saga any idea for where can i find it?
If it's not here, you could learn how to train a voice yourself or just wait until someone makes it
What's gonna happen to all the models on the weights.gg
My guess is that these voice models will be wiped when their website and database close down, some parts of them may be archived by someone else. 
Already uploaded one to my huggingface lol
Since Weights.com never link voice models to their respective original Hugging Face links like ones in #1175430844685484042, "they just stored model files directly on their database", so it will likely happen.
is it just me or is the RVC GUI actually better than Vonovox? Given the identical settings on both software, Vonovox gives a lower pitch and sounds a bit muffled, ending up a less natural voice
Do you mean the "realtime" mode .py found in mainline RVC GUI? That one is the oldest, older than latest mainline W-Okada version. While few latest "Applio RVC" versions now feature "realtime" mode within the same GUI, other than these two RVC GUIs, there doesn't seem other "RVC fork" that would actually feature the "realtime" mode. 
I'm not sure which fork I'm using. But I believe this is the one
https://huggingface.co/lj1995/VoiceConversionWebUI/tree/main
RVC20240604Nvidia50x0 to be specific
is this the mainline RVC GUI?
This link's author name might look unknown to me at first, but this is the mainline RVC GUI.
The file name "RVC1006Nvidia.7z" is what I remember this should be "mainline RVC GUI" for sure, which it is.
which fork is recommended for RTX50s card?
Still Tg Develop's W-Okada fork and Vonovox.
Thanks!
why dost it not load up when i try opping it pls help
I wonder if this channel should be renamed to not ai help channel? @covert lake
Because literally everyone thinks it's this channel and not the one called ai help
Does it work on a 4gb gpu
Case and point
I am uploading voices onto Internet Archive.
Don't upload any of mine since mine are already here
Why do people want to upload their voices to Hugging face instead of Internet Archive?
Really simple
What?
Huggingface is simple
Huggingface is too complicated for me.
I do Archive because it's much easier.
I could record exactly how to upload anything in like 3 seconds if you need help ^^
#✨│ai-help is hidden by default, which is the problem I think
Someone should let the mods know
Well, I tried searching for the voices on Huggingface & I can't find it.
They're all here
https://discord.com/channels/1159260121998827560/1175430844685484042
It's hard to find anything on huggingface but uploading is easy
But most of the voices from Weights aren't on voice-models yet.
True
So what?
Upload people's models who aren't users that post in voice models, like me or Vix
But how?
U can use Internet archive I guess
We aren't archiving that much voices enough.
MY 25th tupac model. lol sounds realy good at 20 epochs me singing . has really good potential. trying soemthing new
As long as the model wasn't made on weights it should be saved
That's what I'm using. Much easier for me.
Wait. Models made from Weights can't be saved?
Why?
Models made using weights are always bad compared to models made using applio or mainline rvc
Worse quality
What is your name 😭
So what will happen to my uploaded models on Internet Archive after Weights shuts down?
They won't work anymore?
They'll stay and work as long as you have downloaded them and archived them in a zip file or folder of some type
But my downloaded models from Weights are already a ZIP file
That's ok then
I also archived them on Archive.org.
👍
So are those uploaded voices on Archive.org safe then? They will still work right?
Yup
After the shutdown?
They'll work even after the shutdown because they're files and files cannot be altered by anyone but you after you download them
Anyways, it still feels sad to see Weights go as a result of their greed.
I feel no sadness since replay still works
They removed simple features that should have stayed
Also Ultimate RVC on Huggingface
The only good thing weights did was limiting how many covers u can make at one time but then they undid that limit
Financial difficulties caused Weights to shut down.
They could have given many good reasons to actually buy premium
Like what?
Not entirely sure but, benifits
Benefits like what?
They disliked paying users as much as free users
I don't know I'm not that smart
Also the paywalls didn't work for them.
True
They added gambling to weights which was actually kinda fun
Should've added it way before they ruined themselves
I might as well say that the devs didn't do a great job of running the site.
The whole website went to hell
For being too greedy?
Yea pretty much
They made too many bad decisions and didn't listen to their fanbase
i been using ai for my programs
They could've put the ads before they put the paywalls.
its been a great tool
but its a shitty programmer lol, really difficult to get good code
They tried ads and that didn't work at all
Why?
Or maybe the should have started a merch store!
What would they sell 😭
Hats, T shirts, mugs
i hear a lot about ai being like a theif ?
i don't think so
ai is trained on open source right ?
If it's creating images or videos it's using stolen media yea
Art taken from real artists that didn't want them using it to train their ai models






