🧬│ai-chat

[gentle trench]

Fair

If I was able to test 3.0 and og myself that'd be pretty cool, I need a full idea of how well it can clone a voice compared to previous versions of lc

Btw does 3.0 have a cvec and spin version or just one?

[glad nebula]

3.0 is spinv2

[gentle trench]

Fair

Spinv2 is in applio right? I can't even remember

[glad nebula]

yep

[gentle trench]

That's good if I ever decide to switch to it

[glad nebula]

spinv2 is amazing, i like it

cvec needs a lot of context to provide good results, so its fine for local infer, but suffers in realtime

[gentle trench]

Context how?

[glad nebula]

whats being said

[gentle trench]

Ah

[glad nebula]

contentvec needs to hear long sentences to provide a good result

it doesnt work properly for short words

spin works different so it handles short sentences better

[gentle trench]

How different would you say it is to cvec?

Like huge difference?

[glad nebula]

spinv2 has a more natural articulation

cvec sounds too fake sometimes

it does have weird pronunciation sometimes

but spinv2 for me has been very smooth

[gentle trench]

I remember older rvc before it became good would pronounce words like "Portal" or "purple" stuff like that completely wrong

Impossible for it to even say in realtime

[glad nebula]

yea thats contentvec

ok for local infer, just bad for realtime

relies heavily in long context (the reason why extra exists in w-okada :p)

[gentle trench]

Explains a lot on why I'm never really satisfied with my models since I use cvec

Do any good pretrains without broken discriminators use spin V2?

[glad nebula]

in this server? none

[gentle trench]

Are there any other servers that have pretrains? Only other rvc related server I know of is Vonovox

[glad nebula]

idk

[gentle trench]

Damn

Your new lc 3.0 uses it but not public :p

So probably would have to wait for someone to make something of similar quality

[glad nebula]

yes

[gentle trench]

Btw why is it private just curious, I think klm also went private but I could just be misremembering

[glad nebula]

expensive to make

[gentle trench]

That's fair, would you ever share the pretrain if someone offered money for the files? (I don't plan on doing that)

I don't personally see the point of it just kinda existing if only one person can use it

[glad nebula]

no

[glad nebula]

i mean, you can always train your own pretrain

[gentle trench]

I have 0 idea of how to do that

[glad nebula]

it's super hard i'd say

but not impossible

[gentle trench]

Where do you collect your audio from? Because the only site I know how to get audio from easily is YouTube and the sample quality would be lacking due to compression

[glad nebula]

i used the dataset of legacy core 2.5
ai upscaled version of ljspeech to 48k, and m4singer without bass singers (i deleted them because their voices were too monotone)

[gentle trench]

Are any of those public?

The datasets

[forest crypt]

https://www.openslr.org/94/

[glad nebula]

https://huggingface.co/datasets/alibabasglab/LJSpeech-1.1-48kHz

https://github.com/M4Singer/M4Singer?tab=readme-ov-file

[gentle trench]

Ty both! Btw is training a pretrain different from training just a random voice model?

I need a guide

[glad nebula]

yes, it has to be trained without the og pretrain, and you need to properly preprocess a multispeaker dataset

and aim for 2m steps

9 days of training for me

[gentle trench]

Alr I need one last thing, local applio

And a guide of how to use that as well

[glad nebula]

2m steps = 2 million just in case

technically you could train more than that, since rvc doesnt get overtrained

the pretrain never stops learning

you could ask for help in the rvc-development chat

[gentle trench]

Will do

[gentle trench]

So technically I could just let it train forever

And it would have a chance at possible being super good

[glad nebula]

xD

[digital onyx]

IM DOING SEASON GREETINGS🎁✨ ANYONE WHO NEEDS SOL CAN CONTACT ME ARE YOU INTERESTED

[digital onyx]

Do you need sol

[glad nebula]

no

[gentle trench]

Anyways, I guess today I'll do some research and look through those datasets, anything you'd say to remove? I heard u said to remove the bass singers I think

What do those sound like?

[glad nebula]

they're named bass-1, bass-2 etc, just remove these folders in m4singer

[glad nebula]

but legacy core's dataset is pretty bad, train EARS instead, its made by facebook iirc

and also the og rvc architecture is not that great... mpd cant clean mirroring and the breaths are not realistic

[forest crypt]

The dataset ears has about 1h per speaker, so I guess it's a good start

[gentle trench]

EARS?

[gentle trench]

So remove mdp?

[forest crypt]

https://github.com/facebookresearch/ears_dataset

[glad nebula]

no... honestly if i were you, i'd wait for noobies refinegan pretrain

[gentle trench]

I'll be waiting a long time them ig, I have no idea where the updates on it are if any

Besides straight up asking

[glad nebula]

he post random samples time to time

#🔊│ai-development message

[gentle trench]

No access

[glad nebula]

go to channels and roles and give yourself access to the chat

[gentle trench]

Which chat is it?

[forest crypt]

Hopefully that will work for me

😿

[gentle trench]

Thx

So refingan will be used to create pretrains in the future or just to train models?

Kinda slow rn

[glad nebula]

both

[gentle trench]

Perfect

[edgy flare]

Hi everyone! I'm researching how often modern AI systems lose context or forget important details during longer conversations. If you've run into situations where an AI suddenly forgets something crucial or gives contradictory answers, could you share what happened? Real examples would help a lot.

[gentle trench]

Also quick question, does wokada tg fork have spinv2 compatibility I know for now currently Vonovox doesn't

I may switch to that one if it does once refingan releases

[glad nebula]

nope

dr87 is gonna add spinv2 soon

[gentle trench]

I'll just wait for that then, and the dynamic slots I suggested forever ago

Fully switching once the dynamic slots are there

[quartz needle]

hey

[winged wigeon]

what you guys think of arc raiders voice changer?

https://www.youtube.com/shorts/QNWP4kQu7Mg

[minor blade]

Altho despite all of these issues and other pretrain stuff you found about time after making LCV2.5, at the end of the day i consider it a "decent" improvement over OG pretrain.

so anyways

[analog belfry]

https://youtube.com/shorts/3sxjdpGSYww?si=ADRSnx7kQhfXr5f-

[gentle trench]

fuck is arc raiders?

[gentle trench]

yoo it's Venom

[analog belfry]

hell ya it is!!

[analog belfry]

i love local worms!

[gentle trench]

Hell yea!

[wicked wagon]

https://youtu.be/sq5450PJmRo

[analog belfry]

we got some up coming bangers at 6 and 8 pm tonight check them out https://www.youtube.com/@LethalProtectorsYT

[blazing garnet]

Who is the best maker and designer of girls' voices in a language but an Arab?

[lofty spade]

For this program + gaming is better the RTX 3060 12gb or the RTX 5060 8gb?

[analog belfry]

new video out https://youtu.be/xY1BWcIfQwY?si=errfKfrsR4CdzDPQ

[forest crypt]

In the end I decided to go with the EARS dataset and (hifi-tts)[https://openslr.org/109/], with the latter reduced so that most speakers have only 1 hour of audio, leaving only speaker 92 with 27 hours.

[wild heart]

Hey guys, I saw in a YouTube video that there's an e-girl filter for the voice models. Is it no longer available because I can't see it.

[forest crypt]

https://media.discordapp.net/attachments/1193767463016472597/1193767889719808030/caption-2-2.gif

[gentle trench]

https://tenor.com/view/all-roads-lead-to-rome-clock-ticking-white-rabbit-alice-in-wonderland-gif-11005866172848930328

yucky yucky creature

[solar torrent]

https://tenor.com/view/kurosaki-koyuki-koyuki-blue-archive-bullying-tennis-gif-120117928253145884

For anime voice models, consider try some Blue Archive characters.

[lime burrow]

👋

[fast sequoia]

hello all/ 1st time share video

[solar torrent]

Sharing a random video with no explanation is questionable, as it can count as self-promoting.

[lapis light]

That looks cool. Subbed 👍

[solar torrent]

[lapis light]

My bad 😩

[main nest]

tbf, there are completely valid and unproblematic uses for these kinds of voices too.
And talking in such a demeaning, dehumanising manner to anybody who's just asking a simple question about them seems unnecessarily antagonistic to me.

[gentle trench]

I see no good uses for those particular voices, I've seen too many people come in here to pull some crap like in this image, which is why I always respond to it negatively hope you understand it's not targeted at you in particular

I assume the worst when anything "e-girl/e-boy" is mentioned

I'm sorry for how I acted earlier tho, I should have simply asked if you needed help with the voice changer

[polar flax]

I've ever seen the one "to get some valorant skin from e boys"

[gentle trench]

Do people even play valorant?

I haven't heard anyone talk about that game in a while :p

[polar flax]

some ppl asking for help in this server did so

[gentle trench]

Lol

[polar flax]

some vtubers I've ever seen were also playing that

besides apex

[gentle trench]

I forget about that game

[fluid nymph]

i dont want rbx from e boys but voice trolling is fun IMO

[gentle trench]

As long as it's just having fun, it's fine

Catfishers aren't people 🗣️

[fluid nymph]

not to bring that topic up as disrespect or anything but what ab the guys who catch older ppl? the ones that ARENT staged

if any

[gentle trench]

Like Chris Hanson?

[still sigil]

@fast sequoia no advertising

[fluid nymph]

uh idk

[gentle trench]

Oh wait Chris Hanson stuff is staged

[fluid nymph]

i was lwk here for the same thing im ngl

👀

[gentle trench]

Technically you're in the wrong chat for help, I could help u two here https://discord.com/channels/1159260121998827560/1159290139609137264
as long as you're only using the thing for fun y'know

Nothing yucky

Just ping me in that channel whenever you feel ready to set it up and see if it works all that

[hollow field]

@glad nebula what’s Rifingan

[main nest]

I just assumed it wasn't targeted at me anyway, since I didn't ask the question ;)
I did use one or two myself in the past, though, where I tried to find a voice that didn't sound like a cartoon character, but also not like a straight clone of a well-known celebrity or something. Used them for placeholder-audio in animation tests/mock-up, and for some RPG-session, to have something to play to the players (think like audio-logs, where I wanted a few different voices)
And I don't see anything problematic or "yucky" with either of that ;)

I do understand the general suspicious stance and caution, though - I know that people can be shit, but maybe that still shouldn't be the general blanket assumption. Call that overly optimistic or naive, if you will.

[polar flax]

how they use the model for or their intention could be that "yucky", or that the voices are anonymously labeled as "e-girl" instead of mentioning the name

[gentle trench]

It makes me happy you use it for normal purposes

[tawdry wraith]

is there any model that clone my voice and sings the same lyrics with the voice provided.

[gentle trench]

Models don't clone your voice, you have to train a model from audio of yourself

Then it can sing using your voice

[minor blade]

That's not how RVC works.

[tawdry wraith]

Can u share a tutorial or help me on it. Will really appreciate it

[tawdry wraith]

How then? Little help pls

[glad nebula]

uhh models do clone voices, thats literally what rvc is about

its even in the name retrieval based voice-conversion

[gentle trench]

Technically the way he asked didn't makes sense in how rvc works

[glad nebula]

ah yea lol its not zero-shot

[gentle trench]

Currently I can't help now until 3 CST as I'm heading into work

[violet tangle]

anyone can help? mine isnt working

[limber nexus]

please ask in #✨│ai-help and provide more info, which voice changer, whats your GPU, did you follow a youtube tutorial

[vapid geyser]

where do i download the ai voice changer

[lean canyon]

Anyone knows a good Ai to create a Model for the best price?

[valid sable]

Anyone knoe how to get permission to post in project showcase?

[deep sluice]

you just post

[elder willow]

can someone refer me to a voice changer? ik there arew good ones here

[gentle trench]

https://discord.com/channels/1159260121998827560/1159290139609137264

[forest crypt]

Maybe you forgot to add the proper tags to the post.

[valid sable]

It says

You dont have permission to post in this

[empty hinge]

Why is it bad when saying 'sh'

[solar torrent]

For help about W-Okada, TTS or any RVC fork, better ask in #✨│ai-help and #1192011222023950368. Because when you ask like this, it would likely refer to your pronunciation issue rather than any program you expected one to answer.

Sending a random YouTube link with no such explanation is questionable.

[solar torrent]

For help about W-Okada "realtime voice changer", you better ask in #✨│ai-help.

[kindred mica]

what are ur thoughts on AI humanizers? a lot of them dont actually work and like if they do they change the entire meaning of the text and its tone

[violet tangle]

i downloaded, right after i deleted it, my mic started going static why is it happening?

[gentle trench]

We love having no context on what you're talking about

[violet tangle]

idk how to explain it but, i did what was in the tutorial on yt, and right after i deleted everything, my mic started going static on all applications

[gentle trench]

Because you download something old and outdated

For more help for wokada ask here btw
https://discord.com/channels/1159260121998827560/1159290139609137264

[simple gulch]

Someone give me a grandma voice model please

Someone give me a grandma voice model please

Someone give me a grandma voice model please

[gentle trench]

https://cdn.discordapp.com/attachments/1237807523986411573/1272638016552173599/attachment.gif

[simple gulch]

fuck yall then

singing and shit

[gentle trench]

[limber nexus]

#1175430844685484042 #🔍│find-models Have fun :]

[faint nymph]

Is there anyone who can help?

[simple gulch]

Ok?? Im asking the best grandma voice im not asking how to find a model

[limber nexus]

well im pretty sure none of us keep track of that here, so the only way to find out would be trying some on your own, because a model can have different results based on your voice

[simple gulch]

Ok but recommend me some grandma voices

Or female middle aged voice

[limber nexus]

i don't know any names of the models

[polar flax]

as said above we ourselves don't recommend particular models as the listed models are made by different members, so again try by yourself

[gentle trench]

With what, btw help is here
https://discord.com/channels/1159260121998827560/1159290139609137264

[neat prism]

HI, can anyone help me?

I'm looking for singer models, but not artists or celebrities, rather singers who have trained their voices and have good quality, similar to the AUDIMEE models.

[gentle trench]

what is audimee?

never heard of it

[neat prism]

Audimee, Convert your voices with their royalty-free voices, you can also train your own voices, and create royalty-free cover voices.

[solar torrent]

Fuck you too.

[solar torrent]

Most voice models in #1175430844685484042 are mostly RVC (Retrieval-based Voice Conversion). I'm not sure what the Audimee even is, but I'm more familar with RVC, which itself is a free, open source program and is used by most members here.

[polar flax]

can the models be downloaded locally and used in rvc applications?

if not, it's just another paywalled garbage

[neat prism]

Yes, I tried to find some models that are not singing artists. On Audimee's official website, they have models of singers that were recorded in studios, which is why I am asking for help from someone who has already trained models in good quality.

[neat prism]

Yes, that one.

Have any of you tried models who are singers but not artists?

[solar torrent]

The "Audi" I'm also more familar with is the car brand. https://upload.wikimedia.org/wikipedia/commons/thumb/3/35/Audi_A4_B9_sedans_(FL)_1X7A2441.jpg/512px-Audi_A4_B9_sedans_(FL)_1X7A2441.jpg

[gentle trench]

that site sucks

[forest crypt]

Well, technically it seems to use RVC, so you should expect to find the ‘same’ quality here — unless, of course, they’ve done their own research and that’s the reason why their models sound ‘professional.’

[solar torrent]

Wow, that's expected.

[solar torrent]

For help about W-Okada, RVC fork or any AI program, you would ask in #✨│ai-help or #1192011222023950368.

[forest crypt]

I’ve tested some models, but they aren’t English-speech models,

https://tenor.com/view/cat-meme-cat-microphone-microfone-gif-2557129155223957570

[solar torrent]

[neat prism]

Do you mean that an English model would affect the audio of a Spanish singer?

[neat prism]

And in what languages did you find it?

[forest crypt]

Well, most of them were in Spanish. Those models belonged to a user who contacted them personally and sometimes obtained studio audio.

Those models no longer exist now because he deleted them

[neat prism]

Nooooo, just what I'm looking for right now 🙁

[forest crypt]

You can achieve the same results by training your own model—just gather a good dataset.”

[polar flax]

using studio quality audio datasets doesn't always mean the trained models are quite flawless (yea including amount of noise & robotic sounds, how the datasets are prepared, and what kind of pretrain & discriminator they were using)

in the end, if theres nothing convincing it's outright scam

[faint nymph]

Google colab voice changer is good ¿

[elder willow]

Who uses Google colab nowadays

[faint nymph]

What do you use?

I am using the paid version

[forest crypt]

I think more depends on your internet connection, iIf you use a paid version, there is no doubt that it can work better for you, but I guess somehow there will always be some latency

[noble halo]

how to use your voice to make the charaters say what you want but it is in their voice

have a vid example to show and can send you

64gb ram 5070ti

basically voice clone and the charatcre voice says what you want them to say

[minor blade]

That sounds like TTS.

[noble halo]

OK, how do I like train voices like I need to take video clips and then extract them using vocal isolation and have long audio voice files

I probably want like a free local tool that does a really good job

Hopefully my specs are good enough

[gritty cape]

im new to this ai server am i welcome guys

[solar torrent]

Welcome, new member.

[polar flax]

I don't see it a decent investment in long term over just buying a decent gpu to run locally

[gentle trench]

Yikes, just use Kaggle, 30 hours for the free version

Don't waste ur money

[faint nymph]

cheap prices for me

[gentle trench]

This is confusing, is there a better way you could explain what you are trying to do

[noble halo]

Yes I have a video link

[gentle trench]

Just buying a 50 series Nvidia GPU for your pc would be a thousand times better for your wallet

One time expensive payment vs your constant small payments to google

[faint nymph]

I live in Turkey, it is more expensive to buy

[gentle trench]

That's fair

[noble halo]

https://www.tiktok.com/t/ZTrdk9Gw3/

[noble halo]

Is 5070ti good

[gentle trench]

It's really good yea

I have one

[gentle trench]

The dude in this video is just my type 😞

Anyways, I didn't see anything ai voice related on that tik tok

[gritty root]

nah

[gentle trench]

Bro has a NASA computer ↑

[gentle trench]

Apparently I was looking at the wrong video

[noble halo]

Maybe this one would be better

https://www.tiktok.com/@sb_tk_/video/7569740646987107617?_r=1&_t=ZG-91BUd9P8HKK

[gentle trench]

I think I get what you mean

You probably want to use applio because tts wouldn't sound good for that

[noble halo]

What guy 😭

[gentle trench]

I definitely got sent to the wrong video

Lmao

[gentle trench]

This one
https://www.tiktok.com/t/ZP8ULQ1Yw/

[gentle trench]

But anyway you wanna make something like this right?

https://www.instagram.com/reel/DKYrdlCAx33/?igsh=MThvdHl1b2tzNG9mYg==

your_local_wormm

I wonder if he'd like this

Instagram

[noble halo]

Yea

https://www.tiktok.com/t/ZTrdytVfB/

Are you a guy or girl 😭

[gentle trench]

Yes

[forest crypt]

That doesn't answer the question

[noble halo]

😭

[noble halo]

Like this

[gentle trench]

Yea you'd wanna use applio and just record whatever you want them to say

Then use models of whatever characters you're looking for

Like say I wanted to do a Scooby Doo edit you'd just look for voice models of them

Then record whatever over a clip of the characters talking

And use a video editor

Mix it all together

[noble halo]

Uhhhh ok

😭

[gentle trench]

Guess

[noble halo]

Guy?

[noble halo]

Idk what applio is yet how to use or install it 😭 it’s so complicated I need a video tutorial

And problem is I’d make my own models bc they wouldn’t have them

And idk how to do that 😭

[gentle trench]

Yea

[gentle trench]

It's actually pretty easy to learn, all u need to do is download audio off yt of a character's voice lines from like a game or something and if they're only in a movie or show u have to download that and remove the bg audio and stuff then train it and hope it turns out well

[noble halo]

👁️ 👁️

What’s good program to remove background noise

[peak merlin]

Which like voice changer does everyone use for the models ?

bc i dont have one rn

[noble halo]

Gotta train a model and then use one or find one somebody already used

[gentle trench]

Depends on ur gpu

[peak merlin]

i need a good free like voice changer like app to use the models

[peak merlin]

wym?

[gentle trench]

What gpu does your PC have

[noble halo]

Needs to be powerful

[noble halo]

Voice mod?

[noble halo]

Dm me?

[gentle trench]

there's 3 voice changers and they all require you have at least a 20 series Nvidia card for ai voices

[peak merlin]

its either a 3070ti or a 3080 ive completely forgot

[gentle trench]

Sure, when I get outta bed and get on my pc I'll help u learn

[gentle trench]

Both of those will work, for help on the voice changer I'll help ya here
https://discord.com/channels/1159260121998827560/1159290139609137264

[noble halo]

Tysm bro

[peak merlin]

alr

[gentle trench]

Voicemod is completely different to wokada, voicemod just has voice filters that puts effects on your voice, wokada puts AI voice filters over your voice that aren't Doo Doo like voicemods' ones

[frank lake]

hiii

[noble halo]

Mentally drained is acc so me 😭

[gentle trench]

Hai

[low palm]

hello is there any alternative for RVC for folks with an integrated gpu?

[covert lake]

replied in #✨│ai-help

[mystic island]

Is there a free ai thats good at naturally removing objects

[marsh lagoon]

Is Applio technologically slow to develop, or is there little room for further development?

[polar flax]

what would you expect it to be then?

[broken ice]

where do I find e-woman

[limber nexus]

we don't have E-Women here

unless you mean voice models

[solid beacon]

@gentle trench hey which ai voice generator should I use for youtube

@limber nexus yeah can you tell me about voice models

[polar flax]

don't ping particular persons

[limber nexus]

please ask in #✨│ai-help

with your question

[solid beacon]

Ok

[delicate shore]

Hi guys

pls help me

[brazen juniper]

@delicate shore

5 Minute Maintenance Warning

Bot maintenance is starting in 5 minutes.

[delicate shore]

how to install ai voice changer in arch linux?

[solar torrent]

You pinged them as if they both know about RVC and W-Okada.

[solar torrent]

For help about trying to get W-Okada to work on a Linux distro, better go to #✨│ai-help or #1192011222023950368.

[solar torrent]

"Applio RVC fork" is not your typical traditional program/application that needs to be updated every week or month. The latest version of Applio RVC is v3.5.1 and was released like 2 weeks ago, if you were unaware of.

[marsh lagoon]

Is crepe good for training a model with voice files recorded directly by a microphone in a quiet place?

[solar torrent]

Where did you get this information from? The rmvpe is generally more superior to crepe for most parts in RVC, while crepe is older and can be used in a few scenario where the audio would work with crepe; otherwise, rmvpe is still more preferred anyway. To ensure your recorded microphone audio has the least background noise as possible, check out UVR5.

[marsh lagoon]

Thank you!

What's the optimal amount of vocal data for model training? I'm planning to use the KLM4 pretraining model. I currently have about 50 minutes of vocal data.

[solar torrent]

No idea, as I haven't trained any RVC voice model for a very long time. Many people here opt for built-in pretrained models for training in general, but any other custom pretrain outside Applio RVC would be some cool experiment if functionally good.

[glad nebula]

there is a lot that can be improved in applio, but it can be expensive to do so due to the fact pretrains are always needed

i know they're trying to release their new stuff, which at the moment is just mrd + mpd

and refinegan instead of nsf hifigan

[glad nebula]

10 minutes minimum up to 2 hours max
training a dataset over 2 hours doesn't really give any audible improvement

[marsh lagoon]

wow

Thank you!

[glad nebula]

[marsh lagoon]

I haven't seen a pre-trained model using Refinegan, but if there's no such pre-trained model, can't i do it with Refinegan?

[glad nebula]

he is still training his new refinegan pretrain, not done yet

you always need a pretrain to train a rvc model

[glad nebula]

ive heard crepe gives more smooth outputs

some say it sounds more realistic

[marsh lagoon]

oh

[glad nebula]

for an experiment try using crepe for training, and for voice conversion use fcpe

[marsh lagoon]

thank you!

[glad nebula]

yea no problem, just experiment with it
if you dont like how it sounds you can always just train a rmvpe model

[marsh lagoon]

👍

What I felt while training the model is that since the method is fine-tuning, the learning is over quickly, and it seems important to continuously refine the dataset by listening to it and change the training conditions in various ways

[abstract sundial]

lil help?

[solar torrent]

Go back to #✨│ai-help for help about W-Okada, and don't try to ask for the same thing in chats, as no one would give you a proper solution here. Hope this clears up.

[covert lake]

well not sure what model does ai overview use, but it needs good prompting

[lime raven]

i wonder why some ai voice models sound like theyve been processed through autotune when they sing lower notes like C2 even though their training set shouldnt make them do that

[lime raven]

this is engineer from tf2

this is something id hear if a voice was transposed down an octave using autotune

[gentle trench]

It's so smart

[covert lake]

they are using smt like gemini 2.5 flash lite latest for the ai overview i guess

[gentle trench]

I actually wondered too what Google used for it

[gusty topaz]

im so confused on how to make this work

[gentle trench]

make what work

[dapper ginkgo]

Hmmm... I left multiple months ago. What did change here?

[forest crypt]

Well, you're still a cat

[dapper ginkgo]

Cats are just peak

[dapper ginkgo]

https://tenor.com/view/spoon-cat-spoon-spoon-cat-cat-apple-apple-cat-gif-254879498273748

NO EMBED PERMS

NOOOO

[limber nexus]

https://tenor.com/view/funny-memes-funny-jokes-anime-anime-girl-discord-gif-5371623980962101408

[dapper ginkgo]

Indeed

no gif perms

[dapper ginkgo]

Which role do I need to get perms to upload images?

I forgot

[forest crypt]

15

[dapper ginkgo]

gosh

did that get increased?

[forest crypt]

[dapper ginkgo]

I could swear it was lower before

[dapper ginkgo]

[forest crypt]

?

[dapper ginkgo]

[dapper ginkgo]

gosh

brain aint braining

[forest crypt]

[gentle trench]

https://tenor.com/view/dancing-cat-cat-kitten-dance-dancing-gif-9607462955328857194

[limber nexus]

[dapper ginkgo]

Hold up... the bot says I am level 23

I want my roles back >:(

i am rank #67 LOL

[gentle trench]

https://cdn.discordapp.com/attachments/1044053532866773052/1253374913453228152/togif.gif

[heavy lily]

HELP ME
im trying to train an ai model off my subtitles for a video and i need help transcribing my videos into txt files

[dapper ginkgo]

Do it with your hands and keyboard

easy

[gentle trench]

so true

[forest crypt]

https://github.com/m-bain/whisperX

[mighty bison]

how do i upload models to kits.ai

[solar torrent]

I don't use Kits.ai as that one has their tiers overpriced than necessary, but there's Weights.com where you don't need to upload an already-existing voice model there. If you mean something else, let me know.

[polar flax]

what level

[full storm]

I mean version

Of gabox

Fv5, fv6, etc

[stray kindle]

ive tried to use gemini 3 pro to work on a video game, its better than previous models but for some reason will commonly fail and very simple tasks, like added a few different items, it will very commonly make up entirely new names for things, forgets core aspects, etc. its pretty unreliable but its still an improvement

[polar flax]

the "highest" doesn't always mean to be the best

[full storm]

which version is the highest tho

[polar flax]

read the complete guide and go figure

https://docs.google.com/document/d/17fjNvJzj8ZGSer7c7OFe_CNfUKbAxEh_OBv94ZdRG5c/edit#heading=h.roiuj54hzww3

[gentle trench]

Fv4 is the best for datasets

Most recent one is fv6

[polar flax]

the latest is some experimental ones https://huggingface.co/GaboxR67/MelBandRoformers/tree/main/melbandroformers/experimental

[gentle trench]

I only use the vocal models of FV

Is there a fv7 out yet

[solar torrent]

The voc_fv7 is currently in beta and experimental, with three beta variants of the same model available in Gabox's Hugging Face repository.

[marsh lagoon]

I'm going to train a singing model, which one is better between legacy core 2.5 and klm4?

[worn fulcrum]

guys can somone help me with this ?

[solar torrent]

For help about W-Okada or Vonovox, go back to #✨│ai-help.

You have been told to go to help channel months eariler, and you somehow did the same mistake.

[crystal vigil]

is this the website or did the app update?

[gentle trench]

Nice to know 👍

[timid meadow]

Yo

When will AGI be here most likely and what will society look like at 2050?

[gentle trench]

Agi?

[timid meadow]

Artificial general intelligence

[gentle trench]

Probably will be better than now

[visual mulch]

hi guys

best voice to girl

to troll

[odd salmon]

straight to the point ✌️

[peak merlin]

can someone help me

for some reason setup and start isnt working

[forest crypt]

-> #✨│ai-help

[gentle trench]

no

[peak merlin]

how do i ge tthe vonovox to work in roblox

[gentle trench]

why do u need it for roblox? 👀

[peak merlin]

trolling my friends dw

how to do i stop the voice from like lagging and sounding buggy

[crimson mason]

what does trancate do in the advanced settings, and what would be a good value?

[humble lava]

Ai Automations someone?

[hollow field]

whats the best local applio

[subtle parcel]

yo does anyone know why my voice ai keeps glitching out with some words like u cant understand them at all

[crisp shale]

If you know u know 😏

[still sigil]

people like you are the reason this exists

[crisp shale]

Bruh I am 21

[polar flax]

it'd be more problematic if it needs taking your selfie photo and yet could be abused by generated photo images

[still sigil]

it is a selfie photo

https://www.instagram.com/reel/DRSyMBqjwPr

ohyhay

Roblox age verification....

Likes

22107

Instagram

[forest crypt]

[polar flax]

in the case of most age verification it's not really an issue as real underages mostly don't know what to do so, but in the roblox's age group system, such predators could generate fake photos for under 18

[minor bear]

hi

[gentle trench]

Alright David bazookie :]

[polar flax]

https://tenor.com/view/sus-cat-cat-stare-stare-bruh-cat-really-n-gif-25597234

[dapper drum]

How about vrchat?

😏

[gentle trench]

I play vrchat :3

[elder willow]

I tricked it with SDXL photorealistic images >:3

[snow geode]

i cant get the voice ai thingy to work

[gentle trench]

What voice AI thing, there's 3

[snow geode]

voice changer master?

idfk the name

[broken raft]

You're talking about Vonovox or W-Okada, correct?

[snow geode]

w okada

its so ass to use it keeps distorting and replaying

trying vonovox

[gentle trench]

Did you get it off YouTube?

[snow geode]

perchance my friend sent it to me

[gentle trench]

Possible sent an outdated version is why it sounded so bad

[gentle trench]

You have either VB cable or line 1 right?

[snow geode]

yurp

[gentle trench]

Nice

[gentle trench]

U should add my vrchat user I play all the time

It's the same as my display name here

[hollow field]

whats the best vocal isolator model
idc about datasets
i just want the best of the best

[elder willow]

hdllo everyone i am a complete biegginer i am looking for a partner with whom i can learn ai together

[gentle trench]

Holy shit the guy from SpiderMan

[ionic sun]

[ionic sun]

wake up

[slim field]

hello guys

[ionic sun]

wake up

[gentle steeple]

Hey everyone
I'm working on a TTS project and found an existing project that does voice cloning from scratch. I'm getting decent results with it, but I have a few questions:

Coqui vs Custom Project: Is Coqui better than building voice cloning from scratch? What are the main advantages?

VRAM Usage: My current project uses all 4GB of my VRAM. Does Coqui have similar VRAM requirements, or is it more optimized?

Speed Issue: My TTS takes about 10 seconds to generate speech. I need it to be under 1 second. Any tips on how to optimize for faster inference?

[chilly lake]

Coqui has been dead for like 2 years already

their package has a bunch of different tts engines, but xtts v2 was the best one, unfortunately the sound quality is lacking

[queen valve]

Hello guys, a quick question... What version is better? Is it MMVC v.1.5.3.18a or VCClient 2.0.78

[heady helm]

what fork should i use deitries of tg???

[gentle trench]

Those are super duper old

For help with wokada go here ^^
https://discord.com/channels/1159260121998827560/1159290139609137264

[queen valve]

Thanks a lot.

[thin bone]

Hey guys

Can anyone help me with something AI related?

[limber nexus]

that's why we have #✨│ai-help ask there please

[gentle trench]

I don't think bro knows what help vs chat is

-# a lot don't actually

[limber nexus]

i'll take this over someone going into chat and flaming everyone

[thin bone]

Hey guys, anyone know a great Uncensored/No restriction GPT model that can seearch through the internet as well? I don't have an issue if it's paid

[thorn night]

[I NEED HELP]

Hi everyone, I have a question and need help:
I'm building a tool that can:
•⁠ ⁠generate script for videos up to 2,5 hours long for many different categorie
•⁠ ⁠translate the generated script to up to 30 languages
•⁠ ⁠scrape articles for script generation
•⁠ ⁠generate prompt for image generation
•⁠ ⁠generate title, description and SEO tags
•⁠ ⁠generate the voice over with Elevenlabs (the entire script at once!!)
•⁠ ⁠connect YT channel for statistics
and many other functions that I'm building.
The question that I have:
I was thinking about charging a fix price starting with 15$ per month, but the user connects his OpenAI and Elevenlabs API Keys instead of working with credits.
What do you think? you have your keys and manage your own spending or rather pay a bit more with credits?
Thank you in advance!

[lone hearth]

If you want to go for a more general audience = Make your credits system

If you need a minimum of tech knowledge = Make people use their own api keys

[thorn night]

Thanks for the support! I've been thinking about this lately... What's the biggest audience for this type of tool:

More experienced people who prefer to use their own APIs, or people who just want the problem solved, you know?

I created it for my personal use and I don't understand anything about launching this kind of product.

[kindred wedge]

Hey folks, I got an AI product idea that I'd love to validate, is that something that's allowed here? @mods

[wide coral]

y

[gentle trench]

That's the second to last letter in the alphabet

[wide coral]

where can i make my own ai voice

[gentle trench]

Applio

[wide coral]

i tried on RVC WebUI but it didnt give me an index file to add to my w-okada

[wide coral]

ty

[gentle trench]

You're welcome btw I will @ you in AI help

[scarlet mesa]

any reccs for unbiased ai?

[ivory niche]

No, that's "&."

[gentle trench]

true true

[mint fox]

hello anyone know any good ai's for trascribing audio to text, and iim talking about a recorded audio wav file

[gentle trench]

https://cdn.discordapp.com/attachments/1044053532866773052/1253374913453228152/togif.gif

who is bro

[timid meadow]

AGI when

[solar torrent]

I'm more familar with the original AGP PC card, the card that came before PCIe. https://upload.wikimedia.org/wikipedia/commons/thumb/6/62/AGP-Video-Card.jpg/960px-AGP-Video-Card.jpg?20100911200856

[solar torrent]

YouTube can also transcript the vocal audio into text, known as auto-generated captions, but as a subtitle and it has to be a video. Some AI websites also offer similar tools, but some often locked behind subscription/payment and you'd pay them to access much more features.

[pliant skiff]

how much vram is needed for flux 2?

does anyone have any idea?

[red epoch]

Guys, is a 4 GB video card for the RX580 Voischanger normal?

[solar torrent]

There are two variants of AMD Radeon RX 580; 4GB and 8GB. The 8GB one is a bit superior than 4GB one, both would work with W-Okada, but the 4GB one would likely struggle with other graphic games. The similar issue also occurs in some mid-tier NVIDIA GeForce GTX 10 series GPUs.

When you try to run the "original" version you got from official W-Okada voice changer GitHub or those tutorials from YouTube, the program might not work but would instead run on your PC CPU. For more help about W-Okada, you can go to #✨│ai-help or #1192011222023950368.

[covert lake]

It can go from 90GB Vram to as little as 8GB depending on how you set it up, I would suggest checking https://huggingface.co/blog/flux-2, some flux.2-dev gguf guide and #1192011222023950368

[rancid shoal]

Hello everyone. My name is Mateusz Wicha and I'm from the United States. I'm excited to join this community. I work in software development and am looking for a partner with exceptional technical skills and abilities. Please connect with me on LinkedIn: https://www.linkedin.com/in/mateusz-wicha-828a45391

[urban torrent]

Hello

[final parcel]

Hello guys

Is they anyone into visual creating here

[limber nexus]

everything i make is visual, you need eyes to see it

[gentle trench]

so true

[final parcel]

Ok DM me

[gray flare]

2027

[glad nebula]

@dawn temple model maker bot broke again, doesnt accept huggingface links 💔

[dawn temple]

Can you please send me the link used?

So that i can check

[glad nebula]

https://huggingface.co/lyery/legacy-core-2.5/tree/main

[dawn temple]

Once I’m on my laptop

[dawn temple]

Alright! Thanks

I’ll check it soon

[glad nebula]

yay

[celest panther]

i need help

when i open the start it just opens a black cmd and doesnt open or nun

[still sigil]

this is an ai server, not a crypto server

[gray flare]

And crypto with AI ?

[still sigil]

nah that was a scammer

[gray flare]

Joke, when I listen crypto without bitcoin = scam

[still sigil]

the normal "click the link i send in dms to open a ticket" and then they ask for your 12/24 word phrase

[gentle trench]

https://cdn.discordapp.com/attachments/1364974592531365899/1371614878716137482/image0.gif

[polar flax]

crypto is actually dead to us people but still used by scammers/frauds

at the end it is mostly ponzi scheme in any form

including those using automation/AI slop

[tacit widget]

Is the #1 site for finding models still voice-models or?

@stark bloom

[gentle trench]

Not a site but it's moderated to only have good stuff in this discord yuh

[amber imp]

but where did it get released for free? i'm so curious about this. i run VC on a pc and a random 'live deepfake' app on my main pc so it wouldn't stress much. i'm really looking forward to see that facefx app.

[gentle trench]

I'm afraid of what you use that for

[broken copper]

what kind of ai tts tools could i use on a 1060 3gb?

[modern wadi]

i am using the voice changer but theres a problem i am facing
i geting the voice very late and too very bad
even at 128

[gray flare]

Hold on there, my friend, crypto has never been this strong, but I'm talking about real-world use, not investment

[polar flax]

in what usage? gpu mining is already dead since 3 years ago, N FTs have been nothing but a worthless joke, the bi tcoin value may keep increasing from last years but fluctuates monthly/less as usual. I think the one that still benefits from crypto is the scammers/frauds as I said.

[gray flare]

Do you know what a crypto is used for?

I'm don't saying about investment bro 😭

[polar flax]

I don't quite understand what you're talking about the non-investing usages

pay 2 earn games? that has been a joke nowadays imo, if not just media for some strange coins

[gray flare]

No way, nowadays companies are increasingly to evading taxes, it's beautiful to see

[polar flax]

that's also what many frauds/money launderers/illegal importers are doing.
no wonder there are many ppl & entities trying to import banned Nvidia gpus (H100, etc.) to China (despite homegrown ones being encouraged to use), probably using crypto in some black market and trying to evade customs and use some different transits.

still it won't benefit us folks at all and solve such problems as poverty, unemployment, etc.

https://tenor.com/view/mad-no-money-broke-gif-12333690400092991540

[icy cypress]

Hi! What is the best ai model at the moment?
Talking about programming and researching

[pliant skiff]

thank you!!!

[solar torrent]

I also told you in #✦│chat message, although I doubt if you have seen that one too.

[glad nebula]

@covert lake who accepts the model maker submissions now? i dont see qc so now mods do it?

[earnest dragon]

I did make a better one that works on all gpus except 50 series just not 50 series because never had a access to a 50 series windows vm to fix it

[polar flax]

isn't it more of the torch and cuda version being used?

[vagrant quiver]

Please advise whether I need a web service API, git, or workflow (Comfy UI). I upload my voice and a reference song, and it should sing along with the vocals, but in most cases it mumbles the words.

[cold fulcrum]

how do i make my voice changer work

[covert lake]

Iirc it was going to be a community vote, would be better to ask @dawn temple since it was his bot

[covert lake]

Please elaborate in #1192011222023950368

[covert lake]

elaborate more in #1192011222023950368

[gentle trench]

U uh, can't advertise

Or smth

[hushed lotus]

is there adolf hitler voice

[gentle trench]

probably

[tacit widget]

found the neo nazi

[gentle trench]

bomb him?

[dawn temple]

Yeah, I’ve been very busy so couldn’t find time to update the bot. It’s not very important for now either as we have zero pending submissions

[glad nebula]

can u give me the model maker role again? or do i need to submit a model lol, i mean, i technically did submit one #1408339243948052501 message

[dawn temple]

Ohh you don’t need to submit it again

There you go

[glad nebula]

yay thanks

[dawn temple]

[spring loom]

What is the best way to do a tts with a realistic sounding voice locally
I was thinking doing a very quick tts model that has personality, and then parsing that through something like vonovox

Or is there a specific tool that would be good for that

[spring loom]

something live*

minimal delay

[blissful blade]

hello this is aryan

i have a question , can anyone help me ?

[hybrid temple]

whats the best ragebait voice model

[solar torrent]

The only best ragebait model is you.

[solar torrent]

For help about, there are #✨│ai-help and #1192011222023950368.

[polar flax]

https://tenor.com/view/bait-cat-is-this-bait-cat-is-this-ragebait-gif-10902073105296387701

[solar torrent]

The quick question won't always come with a quick solution, especially in non-help channel like chat. For help about Vonovox or W-Okada, you can ask in #✨│ai-help.

[hybrid temple]

wait thats a good idea

i could just use my own voice and just troll

[solar torrent]

[placid vigil]

how will vonovox even do tts?

[elder willow]

Guys need your opinion on this random channel on Youtube i found today {kza.media} they post Ai related stuff🤣
🤣.Do you really think it's true. I dont know which world they live in where AI does it all🤣
😂

[forest crypt]

No, ty

[gentle trench]

smartest member ever?

[limber nexus]

Possibly, this is a level of genuis ive never seen before

[marble fractal]

do these vc even work

[limber nexus]

the voice changers? of course they do

[marble fractal]

aight i wanna troll some ppl is ts hard to do?

[limber nexus]

No, unless you've never touched a computer before it should be mostly straight forward

[marble fractal]

gulp

[limber nexus]

oh no...

[marble fractal]

early chrismass present...

[limber nexus]

huh?

[marble fractal]

my pc

was a chrismas present

[limber nexus]

I see, well if you would like me to help you we need to go to #✨│ai-help

[marble fractal]

ooo bet

[teal mortar]

join vc

:C

[pearl blaze]

#✨│ai-help

[gentle trench]

https://tenor.com/view/white-monster-wmster-monster-energy-monster-energy-drink-monkey-gif-15628061433413475006

[solar torrent]

The gooning gooner.

[solar torrent]

https://tenor.com/view/vro-pmo-ngl-gif-10268498521266426722

[limber nexus]

[faint void]

z image is crazy

anyone know if its possible to convert it to image to image

[flint skiff]

Guys does anyone know get rid of deley on the voicve changer

[finite cave]

buy a time machine

[gentle trench]

This is a really good idea

[restive hamlet]

what are some good, easy, and stable notebooks for rvc voice cloning?

[gentle trench]

applio

[minor blade]

The only one is applio

[wind knoll]

how do i make the voice changer work on valorant???

i am using the cable input abd ouput cb audio virtual cable#

[gentle trench]

Try turning the graphics down, also what gpu do u have

[wind knoll]

radeon rx 570

work on discord tho

[gentle trench]

O

Hmm

Yea just try turning down game graphics

[wind knoll]

i did no working 😔

[gentle trench]

😔

Might be time to upgrade your pc

Get Nvidia GPU it runs ai way better than amd

Around 3060 or higher is best

[wind knoll]

yeah but its odd that on discord it works

but not valorant

[gentle trench]

Games take a lot more GPU to run and with the voice changer also taking GPU it stutters and might not work at all

It's weird

[wind knoll]

😞

[granite topaz]

Anyone know what software / system was used for this type?

https://www.youtube.com/watch?v=TjPa9Hq-UCM

[wind knoll]

the problem was my anti virus 💔

[restive hamlet]

oh wait yes applio's actually going

we're training a model with cloud based applio let's goooooooo

[gentle trench]

ye applio has cloud'

kaggle is the best option

30 hours a wekk for free users

[restive hamlet]

success

i knew cloud-based applio training would pay off very well

[gentle trench]

very much

[maiden quartz]

it ain't loading the voice models

[maiden quartz]

guys i got a problem i got everything set but i can't hear the voice

[shell pilot]

Does any one have the app from Socksapplo

[gentle trench]

what even is that

[shell pilot]

hes a youtuber

he posted a video where he showed how he made ai Minecraft villagers talk but i cant find his video any more

[gentle trench]

oh he just used applio and converted random stuff he said with an ai villager voice

[shell pilot]

how do i use applio?

[gentle trench]

-rvc

[pine acorn]

📚 RVC Documentations

AI HUB Docs

https://docs.aihub.gg

🍏 Applio Docs

https://docs.applio.org

[gentle trench]

here's the guide ^^

[still crescent]

why ru a worm

[gentle trench]

Because I wiggle and stuff

[still crescent]

thats lwk tuff

[gentle trench]

Real

[still crescent]

yes

[dusty moon]

hi

Do you know where I can buy a good ai girl's voice?

https://tenor.com/view/middle-finger-foot-gif-19157514

[gentle trench]

Ew...

[lime lodge]

hey guys, is there an open source/freeware tool for midi/text to ai vocals, or maybe rough vocals+midi+text to ai vocals?

[lapis edge]

So-vits-svc is basically abandonedware even the fork verison

[latent sky]

I have a 16 gb ram, 4 gigs rtx 3060 and i9 11900h laptop, I want a realtime voice changer, for games like valorant or minecraft or sm

[faint void]

🤖 Beep 🤖 Bop 🤖 Bip 🤖 Generating voice changer.......... 🔃

Voice changer successfully generated ✅

[latent sky]

[latent sky]

damn

thanjs

i owe you one

[wraith rapids]

Anyone

How to change my voice into girl voice using Voice changer, I'll play $

[ripe moth]

guys im using the real time voice changer but the voice isnt changing

am i doing sumn wrong?

[faint moss]

mercor ai

[cedar field]

helllo bros

[faint moss]

hi

[faint void]

dm me

[minor blade]

Just read the docs and/or go to the #✨│ai-help

[latent sky]

which docs??

[minor blade]

https://docs.aihub.gg/realtime-voice-changer/local/vonovox/

[latent sky]

thx

I have a 16 gb ram, 4 gigs rtx 3060 and i9 11900h laptop, so will it work?

[minor blade]

@wraith rapids Read this and install Vonovox.

[minor blade]

It should be enough.

[latent sky]

oh okok

thankss

[gentle trench]

https://discord.com/channels/1159260121998827560/1159290139609137264

[gentle trench]

Yuck

[cedar field]

Hellloo

Lately, I've been mixing Afro and Brazilian Phonk, I really love these rhythms. I posted a track in the same style on Spotify. I'd love to hear your thoughts 🙂

[gentle trench]

this isn't the chat for music but sure

[forest crypt]

https://cdn.discordapp.com/emojis/1422182881626816585.gif?size=48&animated=true&name=laughandcry

[gentle trench]

https://tenor.com/view/no-cyn-sweetie-meme-murder-drones-gif-15307302106880553525

[idle mica]

Can someone upload TheBurntPeanut's voice model please 🥺 🙏 💯

[gentle trench]

Who is that

[hollow field]

@hollow tangle we should be able to use our own models for UVR UI

that would be cool

[limber nexus]

https://tenor.com/view/nuh-uh-punch-gif-14008503396845178733

[high thistle]

Anyone got issues with chatgpt for last few hours?

[main sluice]

how do i use the voice model like how do i download it

[solar torrent]

I use Google Gemini.

[solar torrent]

To request a voice model to be uploaded into #1175430844685484042, see #1159289738314919936.

[vale walrus]

Where do y'all make the covers?

I use weights.gg but is there something better?

[solar torrent]

Applio RVC fork is the only option where you can make AI cover and even train a voice model, either run on your PC or online, although this one can be more complex to set up than Weights.com.

[glass gate]

hi

everyone

i need help

acctuly i am working on project called Wibblio

[indigo sluice]

Dear community!

I am working on a research project dedicated to identifying fabric types using artificial intelligence. To train the model, I need to collect a high-quality and diverse dataset of images. I would be very grateful if you could join and upload photos of your items.

How to do it correctly:

Folder link: https://drive.google.com/drive/folders/1WrWgtRebIOpOBBi3Fqw5gwX_NlmikxLf?usp=drive_link (All photos will be checked manually and i can delete every photo, so pls no shitty photos or memes)
Structure: The folder already has categories (Cotton, Denim, Wool, Polyester, Silk, Linen, Leather). Please upload photos to the appropriate subfolder.

Photo requirements:
Take photos as close as possible (macro) so that the weave of the threads and texture can be clearly seen.
Be sure to use a flash or flashlight for good lighting.
There are some examples from me.

Important: Upload photos only of items that you are 100% sure of (check the tag/label before taking a photo).

Your contribution is very important for the success of this project. Thank you for your support!

[limber nexus]

Just, try to make it fit in as best as you can

[zenith plover]

i just saw im int he wrong chat

im blind

[limber nexus]

it's fine

don't worry

[gentle trench]

https://discord.com/channels/1159260121998827560/1159290139609137264

[latent sky]

yo I need help downloading wokada from starting idk sht

[calm path]

@limber nexus are you able to help with an issue im having with wokada?

[limber nexus]

please read the help guidelines before asking and i may be able to yes

[calm path]

i actually just managed to fix it, sorry to bother

[limber nexus]

No worries, glad you got it fixed

[zenith plover]

is providing a link alongside the message on help ai chat allowed to enhance the context?

[solar torrent]

If you have any additional issue when running W-Okada or Vonovox, you can go to #✨│ai-help or #1192011222023950368.

[solar torrent]

Does it? The only link that's understandable would be a tutorial video or a guide link you got. With an image or screenshot would be better.

[zenith plover]

just a video of me showing the progress of something im doing for context.

[dusk niche]

oh i apologize for asking in wrong channel

[solar torrent]

[zenith plover]

yuuka I didn't get any response.

[solar torrent]

https://tenor.com/view/blue-archive-noa-ushio-noah-nervous-acomu414-gif-431796544209125674

[zenith plover]

if its still either reasonable or not

if not i won't do anything

[solar torrent]

You can also send a video screen record of your monitor screen.

[minor blade]

@forest crypt Are you there?

[forest crypt]

?

[minor blade]

Oh, it was just something related to localtunnel + gradio on kaggle. It was behaving even more slower than usual.

I decided to just switch temporarily back to ngrok

[forest crypt]

Check the localtunnel status page

[forest crypt]

Ngrok is fine, only there's a bug that's reported very rarely

[full olive]

hey can anyone help me with this issue im having? i downloaded the ai voice changer and a while ago it worked perfectly but now when i try to run it the start_http won’t run, it was working a few minutes ago until i restarted my pc

[gentle trench]

Bro doesn't know they download an old voice changer

[unique ember]

ai stack devs

dude ai is doing it

ur not in the process

[gentle trench]

What gpu do u have, start http is for an old version of wokada

[full olive]

nvidia 3050

i used it almost everyday a month ago and today i tried using it again and it just wont run

[gentle trench]

You should get Vonovox, the current beta is kinda messed up so I'll give u the version right before it

-rt

[full olive]

where do i download that

[pine acorn]

💻 Local Realtime RVC

Guides for Programs that use RVC Models in Realtime for Calls/Games

• Vonovox

A Realtime Voice Changer with similar performance to Wokada Deiteris Fork, with extra features, but supported only for Nvidia GPUs on Windows. and without cloud options GUIDE

• Wokada Tg-Develop Fork

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

• Wokada Deiteris Fork

Most suggested WebUI with the best general support for many platforms. GUIDE

⚔️ Wokada Deiteris Fork vs Vonovox

For Windows Nvidia, Both Wokada Deiteris 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 Deiteris Fork Pros&Cons & Vonovox Pros&Cons

⛔ Outdated/Discouraged

These options are not recommended for use.

• Original Wokada

Not suggested, older versions in youtube tuts are even way worse. GUIDE

• RVC GUI Mainline Realtime

The program is worse compared to the ones above, and much less updated. GUIDE

[gentle trench]

It's that first link

[full olive]

okay thank you lots!

[gentle trench]

I forgot I wasn't in ai help

Welp too late

Btw @full olive these two downloads are what you will need https://github.com/dr87/Vonovox/archive/refs/tags/v1.6.9.zip

https://software.muzychenko.net/freeware/vac470lite.zip

[full olive]

oo kk

so once i download both files and put them in one folder i just run it?

is there a video i could watch? i feel like i did something wrong

[gentle trench]

U don't put both zips in the same folder, for vac lite run setupx64

After u finish that one for Vonovox run setup, then after run start

[hasty anchor]

do we have any cloud voicechanger?

[gentle trench]

Yes we do

I dunno how to properly use them tho, I'd ask in AI help channel

[full olive]

ahh okay

[clear sparrow]

Does anyone know about any kind of program/job/internship that could feasibly offer tuition assistance/reimbursement for a BS in Applied AI?

[zenith plover]

oh

[teal token]

Yo wsg y'all

[teal token]

👍

[mild escarp]

z

[forest crypt]

x

[clear sparrow]

no way!!

what a good guy

[teal token]

Ohk
But why exactly?

Just wanted to help

[clear sparrow]

lol it’s all good

i’ll gpt it…

another human connection killed from the start…

jk jk

just wanted to explore options for financial assistance with my degree, see what’s out there

[teal token]

Lmao you’re fine. I wasn’t trying to kill the human connection 🙌Just figured the guide might save you some time.

[clear sparrow]

Hold on

Are you for real?

I thought you were a bot or sending spam

[gentle trench]

https://tenor.com/view/teto-kasane-plush-teto-kasane-fact-checked-gif-14951431768440204450

[teal token]

Lol nah, I’m a real person

I just type clearly sometimes 😅

[teal token]

Hey
Wsg

[gentle trench]

haiii

[teal token]

Yo hru doing?

How is your day going?

[gentle trench]

I'm alr, kinda tired tho

[teal token]

Ohk,
You need some sleep

[gentle trench]

nahhhh

hru btw

[teal token]

I'm good
Just chilling on my bed 🛏️

[gentle trench]

[compact spire]

has anyone here used z-image?

[polar flax]

the way you used emojis looks like a clanker

[clear sparrow]

Oh…

What was your message that you responded? Something about willing to help? That would be lovely…

[elder willow]

Chestnuts roasting in the fire

[knotty salmon]

Does anyone know why AI companies offer free plans? Is it really all in hopes that people will like it enough to subscribe?

[teal token]

fix my emoji game then!

[teal token]

Yh,
I said
I put together a really clear guide on this that actually helps. If you want it, just DM me "guide"on FB and I’ll send it right over.

[clear sparrow]

😔

[teal token]

Hey,what's wrong?

[gentle trench]

https://tenor.com/view/battle-droid-compute-think-star-wars-gif-17180363

[compact spire]

hello?

[teal token]

Hi

[compact spire]

Has anyone here used Z-Image?

[white parcel]

It's great, I've only heard good things about it

It's light-weight and can output realistic pictures, including text

And it's turbo version

[compact spire]

Unfortunately they don't have a "official" server

-# and don't tell me about the server on their site, that's a ToS mess

Though thankfully the model is also hosted on Tensor (of all places)

[white parcel]

i thought u wanna run it locally

[solar torrent]

Do you have any idea about what the Z-Image is? Because for image generation, I simply opt for Stable Diffusion and certain related tools.

[polar flax]

"Single-Stream Diffusion Transformer"

[solar torrent]

[compact spire]

eh, i have terrible specs

[compact spire]

cracked up model if i say so myself, its by far the best possible model for realistic gens

followed by Recraft

I'm actually working on an excuse of a project to push certain aspects of my skills to its limits

Something something to create some kind of "lost phone" thingy inspired by certain games, which is why I use realistic models, to create contacts and connections that didn't exist

[marble fractal]

[gentle trench]

Skills?

[solar torrent]

Issue.

[gentle trench]

Real

[teal token]

anyone here been trying to make a bit of side income w/ ai?
i’ve been messing around w a bunch of tools this week and kinda threw together a simple lil guide on what actually works vs what’s just hype.
nothing crazy lol, but if anyone wants it i can drop it.

[gentle trench]

Can't make money off ai here not in this server nuh uhhh :3

[teal token]

Ohk

[gentle trench]

What even is Google antigravity

[restive hamlet]

hi

[sour oak]

hm

[crimson hare]

hi ppl! a quick question. is this server support AI?

[gentle trench]

https://discord.com/channels/1159260121998827560/1159290139609137264

[marble fractal]

NOPE zero ai here! just ignore the ai chat ai images and ai videos and weight models and server name and you’ll be set!

[gentle trench]

https://tenor.com/view/white-monster-wmster-monster-energy-monster-energy-drink-monkey-gif-15628061433413475006

[compact spire]

i do pixelart

[gentle trench]

nice

real art is peak

I made my pfp

[compact spire]

me when i meet people who don't take ai image generators seriously

oh btw z-image can do figurines too

[elder willow]

i can do figurines too

im an artist

but i cant post any images so you just have to trust my word

[gentle trench]

I trust the magicalmagicman

[knotty salmon]

How are people doing tonight?

[soft dawn]

пretty normal, i think

what about u?

[solar torrent]

There are #🎨│art and #🏙│ai-images. What do you mean you can't post an image in this server and wanted other members to "believe" in you?

[elder willow]

my incredible artwork is posted

[solar torrent]

In #🏙│ai-images, this channel is all about "AI art". For arts made by humans, there's #🎨│art.

[elder willow]

yes, i know

[solar torrent]

[elder willow]

[tranquil lantern]

@stark scarab Hey you're the one maintaining UVR colab right?

Is there like a way for the output inference to not automatically turn the volume down

[light spoke]

If i have an account make with full AI model on fanvue, can i put her to show you or this is not for this? I mean the server, sorry for bothering

[elder canopy]

Hey everyone! 👋

I'm looking to try out Riffusion AI but I need an invitation code to get started. Does anyone have a spare invite code they could share?
Would really appreciate any help!
Thanks! 🙏

[gentle trench]

What even is that

[dire thicket]

Sup

[knotty flame]

guys vbcable audio website is not reachable anyone has file?

[worn juniper]

Hi guys. Is there any good text to speech websites?

[icy axle]

Hey, would it be possible to port Yuzu/Ryujinx emulators or at least its available forks to Xbox Series X Dev Mode via using AI and giving it DirectX12 capabilities for Xbox compatibility? I feel like that would be an interesting project.

[covert lake]

I already replied to you in #✨│ai-help

[covert lake]

I don't have the knowledge about building something on the xbox series x dev mode (which seems to require to be an UWP app specifically made for Xbox) or a switch emulator, but I doubt you could vibe code a port for an emulator since it seems something pretty niche and hard and LLMs just predict the next tokens

There was a whole gameboy emulator built by AI for Windows, which wasn't perfect: https://github.com/djkf/vibe-gameboy-emulator
But the gameboy is way older and more explored

I mean you could try or ask such support from forks but I'm not sure, nintendo switch emulation is still developing

[faint atlas]

the VB-Audio cable website is down, does anybody have the download file??? It's been down for a week and i can't find the download, and there's no alternatives that i can understand

[faint atlas]

Tell me if you find one

[marble fractal]

Whats the best real time vc you could possibly get

[gentle trench]

vonovox

[forest crypt]

That question answers itself; in fact, you use it daily to talk to others.

||🗣️ ||

[hoary crane]

hi

[solar torrent]

The only realtime voice changer you can use to talk in voice chat is your real voice. An AI realtime voice changer with a specified voice model could make you look fake. That's all about it.

[marble fractal]

tyy

[gentle trench]

https://cdn.discordapp.com/attachments/1044053532866773052/1253374913453228152/togif.gif

[unreal edge]

I need help with the voice

i turn the voice changer on, my mic is set on the right stuff. but i cant hear it

[solar torrent]

I have a job behind the screen, and I don't seem to need this one.

[coarse valve]

Character.ai used to be fun before the age verification bullsh_t.
Why does anyone want their face or ID info stolen only to access advanced settings?
This has never happened to me before until now. I tried deleting my account, adding it back in, and fixed my birthdate, yet they're still letting us verify their age, even when they're still 18+.

[solar torrent]

So, your statement just says anything about you.

[coarse valve]

I'm actually 18-20+, and I'm kind of pissed that 18+ people who joined character.ai after creating/re-creating their account, or have been on that site for long still have to scan their faces.

[solar torrent]

Nah, you were overreacting to that one. To get help about Character.ai website, you should contact any support who's responsible to the website, not AI Hub by Weights which is where you're here.

[cyan tangle]

how do you change the voicce of a song?

[solar torrent]

Sounds like an AI cover which changes original voice from one to another individual or basically apply some effects to vocal track.

[cyan tangle]

is that the name of the webiste?

[solar torrent]

Do you think "AI cover" is an actual name of a specific website? The term AI cover refers to when you apply an AI-generated vocal over original vocal audio.

[cyan tangle]

i know what it is, im asking where do i find the website for it dufus

[solar torrent]

There's Weights.com.

[compact spire]

Hm

Tensor seems to stay down for a while now

[icy axle]

Antigravity AI now has weekly rate limits, what its the best local AI thats compatible with my ROG Ally for coding?

[full olive]

i download vonovox and it setup everything up already but when i press start it automatically closes why is that

[limber nexus]

please ask in #✨│ai-help and read the help guidelines before asking

[limber nexus]

please ask in #✨│ai-help and read the help guidelines before asking

[sage delta]

Sup guys, where can i find how to create my own voice model?

[gentle trench]

-rvc

[pine acorn]

📚 RVC Documentations

AI HUB Docs

https://docs.aihub.gg

🍏 Applio Docs

https://docs.applio.org

[gentle trench]

the applio guide and ai hub docs should help with that

[sage delta]

thx

[gentle trench]

np ^^

if u need more help tho ask in https://discord.com/channels/1159260121998827560/1159290139609137264

[covert lake]

https://www.eurogamer.net/sega-acknowledges-strong-resistance-as-it-turns-to-ai-but-says-itll-use-technology-appropriately

[gentle trench]

hmmmmm

looks like Nintendo may beat sega as they have said to not be using ai at all

for now

[patent owl]

hi, I installed the voice changer but no sound comes out from the voice changer, someone knows about it ?

[lime patrol]

Can anyone tell me how to stop unchanged voice going through followed by changed voice?

[limber nexus]

i already told you to ask in #✨│ai-help being sure to read the help guidelines before asking

[covert lake]

What? Nintendo Switch 2 performance is all thanks to AI Upscaling

[glad flicker]

How to get andepzai hub back?

[gentle trench]

I didn't know that

Guess they both are stupid lol

[gentle trench]

What even is that

[wide dome]

@hollow tangle uvr5 hugging face down

[odd nacelle]

@finite cedar What mic do you use?

[gentle trench]

Lemme guess, a scam?

How original

[limber nexus]

I would've never known

[low pecan]

anything better than rvc ? or this is best? I want to change voice to someone else's voice.