🧬│ai-chat

[bold rampart]

You mean the tool is old already?

[hollow tangle]

ye

[bold rampart]

Oh, I didn't know that, because I just knew about this

[hollow tangle]

its 4 days old

https://tenor.com/view/now-old-man-the-future-is-now-old-man-gif-9677657

[surreal yarrow]

have i ever talked here

[karmic escarp]

https://youtu.be/c9X4fshiu78 made with ai

nvm, it got taken down

[warped pagoda]

video?

or music?

[plain cove]

ooh i have same meme but other text xdd

[stiff drum]

so ive been working on integrating an open webUI bot into a discord bot but every time i try to run the install command it gives me an enoenet error. no clue whats causing it. the log files say it doesnt see the json even though the json is there, unmoved, and unchanged

does any body have any ideas?

[stiff drum]

well, i fixed it. now im getting a sharding error. YAY

[covert lake]

This is so real ngl

[kindred kelp]

Lol

[stiff drum]

that describes it perfectly

[minor blade]

#1159290752195633273

[junior minnow]

#1175430844685484042

[tacit isle]

release this

[lethal flax]

why can I not make characters on the weights gg app

[sage orchid]

Is there a local method of creating sound effects from a text description?

[chilly lake]

stable audio tools

[dusty atlas]

Generate Image is slow for me

[stark scarab]

My daily life

[remote verge]

hi

[elder willow]

Is there a sort of AI website builder for free?

[native thunder]

Uh

[hoary inlet]

Heyy

Do anyone know how people make captions with anime stickers 🙄

https://youtube.com/shorts/Aom3dijdflM?si=YCv2LJL3dVMBUElZ

[potent iris]

hi

[soft iron]

i'm here to make AI covers of caseoh 😄

[polar flax]

chatgpt, otherwise your prompting skill issue

[covert lake]

Do u need help? If so, u can say ur PC GPU

[twin temple]

Hey guys,where i can find Doofenshmirtz voice?

There is no voice in channel

[covert lake]

You can search rvc ai voice models at:

• #1175430844685484042
• In #🔍│find-models , Do /find with @hidden grotto
• https://weights.gg/ (login required)
• https://huggingface.co/models (but watch out cus in hugging face there arent only rvc ai voice models)
• https://voice-models.com/
• https://thevoicemodels.com/ (for Turkish Models, login required with discord and level 2 on their server)

if there isnt one, you can:

• #1159289738314919936
• #1191429836321849435
• make it yourself with our docs guides https://docs.ai-hub.wtf/essentials/how-to-make-voice-models/

[hidden grotto]

:wave: @covert lake, How can I help?

Available Commands:
• @weights find <query> or /find <query> - Search for RVC Voice Models
• /create - Create an AI Cover
• /image - Generate an Image

[covert lake]

real

[twin temple]

guys why my voice ai laggging?

[minor blade]

https://cdn.discordapp.com/attachments/1159289354439626772/1332508441234767955/sponge-bob-night_.gif

Use deiteris' fork instead, we don't support voice ai

[twin temple]

i use start http is it?

[covert lake]

You're following an old YouTube tut of wokada, don't use that, tell ur PC GPU in #🔍│help-w-okada

[twin temple]

K

[wicked cloak]

has anyone been able to fine tune fish speech locally? the training process throws up a whole bunch of errors about lora keys not being loaded and the resulting models ive tried training aren't able to speak properly

[minor blade]

Yep, also read the guide.

[twin temple]

Guys, why does my voice with this voicechanger sound like just a whisper? this didn't happen an hour ago

[covert lake]

Try increasing the out

Else show ur wokada in #🔍│help-w-okada

[mental goblet]

Guys which model will be most similar to voice from changed speech from Death Note. I mean changed voice from L's notebook or video recordered by Kira 1, 2

[minor blade]

You can try searching any Death Note models at:

• #1175430844685484042
• In #🔍│find-models , Do /find with @hidden grotto
• https://weights.gg/ (login required)
• https://huggingface.co/models (but watch out because in Huggingface there aren't just only RVC AI voice models)
• https://voice-models.com/
• https://thevoicemodels.com/ (for Turkish Models, login required with discord and level 2 on their server)

if there isn't one, you can:

• #1159289738314919936
• #1191429836321849435
• Make it yourself with our docs guides https://docs.ai-hub.wtf/essentials/how-to-make-voice-models/

[hidden grotto]

:wave: @minor blade, How can I help?

Available Commands:
• @weights find <query> or /find <query> - Search for RVC Voice Models
• /create - Create an AI Cover
• /image - Generate an Image

[bright dock]

👋🏻 hi

[minor blade]

#1159290752195633273

[ruby onyx]

Question for the ppl here

How are people making these animations of plankton or SpongeBob on TikTok I’m trying to do my own with the songs of my choice

But I can’t find how to make these animations anywhere

[kindred kelp]

Use stable diffusion for videos imo

[surreal kettle]

Hello. can you please give me a link to the github where you can download AI Voice Changer

[lean tendon]

helo

how does Stable Voice 2 compare to the original v2?

[dapper drum]

where can I find the virtual audio cable line 1?

[covert lake]

https://rentry.co/ForkVoiceChangerGuide#virtual-audio-cable vac lite

for help ask in #🔍│help-w-okada

[ebon flare]

What should I do if my microphone is not selected in this "in" field?

[tepid basin]

Does your mic work in Discord or anything else

[final meteor]

When people type "280 Epoch" in the title, does it mean I have to put 280 somewhere? Literally first time using this vcc, or any for that matter

[chilly lake]

it means how long it was trained

[final meteor]

Are "in" "out" and "mon" important? I don't hear any difference, just the volume

[rare marsh]

hey

whats the new rvc real time voice changer software?

[ionic pumice]

did you mean changing your voice realtime or

cause that's not what rvc stands for

[rare marsh]

i mean real time voice changer

yes

windows

gtx1660 4gb

what

why

[polar flax]

what are you smoking of? there are only 1650 4 GB or 1660 6 GB

[pearl zodiac]

little guide to reasoning models : https://synaptiks.ai/p/from-base-models-to-reasoning-models

[dapper drum]

I love ai voice changer

[solar torrent]

https://cdn.discordapp.com/emojis/1015964749470650400.webp?size=48

[novel acorn]

hi

[tender pier]

what do epochs mean, im new here

[covert lake]

epochs are a unit of measuring the training cycles of the AI model

basically the amount of times the model went over its dataset and learned from it

they don't mean how good is the model, it's just an info provided on how they trained the model by the model maker
More ≠ better
Less ≠ better

There's no way to determinate how good the RVC model is until you try it out or listen to the audio samples if there are

@reef marten poease don’t promote

[covert lake]

use #🔍│help-w-okada

@flint saffron use #1159289738314919936 not #1159516963014451302

[chrome hedge]

Hey, I was wondering if anyone has any information for like Image/video to Video AI rendering software? Like I specifically want one where I can render it myself would be preferable. I'm assuming most/all of the online ones are either bad or paid services.

I'm using AMD CPU and GPU btw. So one that facilitates that field would be nice. Thank you for reading this smol wall.

[covert lake]

like text/image to video ?

[chrome hedge]

Yes

I've been meaning to make some wallpapers animate with some old photos I had, video to video if possible would be interesting too.

[covert lake]

text/image to video AIs:

• Locally (runs on ur pc):
  • pyramid flow (Image/Text to Video)
  • cogvideox 1.5 5b: Image to Video, Text to Video
• Cloud (remote good pc, running on an online website for example, easier to setup):
  • Weights.gg
  • pyramid flow (Image/Text to Video) (HuggingFace Space)
  • OpenAI Sora (paid only, in some countries)
  • lumalabs
  • Hailoua AI

I don’t think there’s any video to video tho

also AMD gpu isn’t the best for ai

[chrome hedge]

Is it just slow rendering or just bad output?

[covert lake]

I meant as compatibility support

you might wanna research if there’s amd support for those AIs

Nvidia is the best for most support bc of their CUDA and TensorRT cores

there’s also Zluda which is a cuda emulator for AMD, might wanna check this

[polar flax]

you can try SD/flux animate though mostly limited its feasibility to just short clips/animated GIFs

[tropic sail]

Hey there. Regarding live voice changer okada is still the best or anything changed ? I ears there is also eleven labs but not sure if it works live. And they probably don’t have so many voices

[solar torrent]

Detris' fork W-Okada works best. The original version has no update as of now.

[tropic sail]

Thanks I will check this out.

[tepid basin]

Red name is crazy

[tropic sail]

oh ? will check this too I guess

[ionic pumice]

that was pointed at namari's role color

[tropic sail]

🤪 oupsie dummy me

Im sleeping awake half dead brain is not braining

[stiff tinsel]

Ever had a customer support call where an AI either completely failed or totally surprised you?

[hollow oasis]

lol nope coz everytime the support calll was picked by a human

[late nest]

Hey everyone, last week we released Spark Engine v1 after over a year of public beta!

It’s the most powerful no-code AI sandbox available right now with over 80 models to generate text, music, images, videos, web search and more.

https://sparkengine.ai/

Just looking for people to let me know what they think at the moment 🙂

[vestal topaz]

Do people still use tortoise tts?

[chilly lake]

tortoise tts is slow and has crappy zero-shot

[elder willow]

ai bad at maths

[glass junco]

when did tupac make this?

[stiff tinsel]

Well things are changing... At least from the way I'm seeing it. Currently I'm working with such a company that allow businesses to build AI voice agents using drag and drop. And the kind of videos I'm seeing on Instagram regarding that is just insane.

[terse quiver]

help me to create a picture of two people holding hands on a table in gold

[hollow oasis]

well yeah that's true, since ai is coming to the world I think we are going to see a lot of changes which we never expected, also what you mentioned here is super interesting. is there anything more to know about them?

[hollow oasis]

if you want something which is hard to understand by ai versions, I suggest you to create LoRa models and train them with the type of images you want to create, sometimes ai can't really understand what you mean since it's not human but when you create a LoRa model and train it with the images u want, it can help a lot.
also you can use good ai models like weights, stable diffusion, kling ai or flux 1 model

[steep pebble]

guys how do I increase ai voice changer's quality?

I tried to change a audio's voice,it works. but kinda low quality

[tepid basin]

You're using the latest Applio?

[steep pebble]

no

i use RVC1006Nvidia

[tepid basin]

Ohoh

I'm dumb you're doing realtime lol

-realtime

[red frost]

Interaction has expired, use the command again for a new interaction.

💻 Local Realtime RVC

• Guide for deiteris' modified W-Okada's Voice Changer Client (Fork)
• Guide for W-Okada's RealTimeVoiceChangerClient
• How to use real time conversion with RVC (voice changer)
• RVC TTS

[tepid basin]

First link on there is best w-okada realtime

Amazing guide

Deiteris fork real nice

[steep pebble]

nah

i want to do not real time

[tepid basin]

Wait I'm double stupid

[steep pebble]

like change a audio file's voice

[tepid basin]

Use Applio then lol

Sorry just woke up

[steep pebble]

is it free?

[tepid basin]

Ye its a fork of what you're already using

More updated more good

https://github.com/IAHispano/Applio/releases

Only other ways to improve quality is
• Make sure your input audio is clean (shitty stem separations can cause some issues. If its clean and well-enunciated, there's not much you can do)
• Use a better model lol

If it voicecracks you can mess with crepe hop-length but thats more advanced. RMVPE is by far the most stable

[steep pebble]

thank you so much

[tepid basin]

Unsure what style of edit you're doing, sometimes it helps to mimic the voice if you're recording it yourself

I usually record like 5 takes of the same vocals and chop in the good ones

[abstract wolf]

no

[terse harness]

ok I have a problem with the software the file called RVC GUI.bat does not launch even after a long wait what should I do???

[median kettle]

Hello friends, I want to add my own voice to a song, how do I do this? Can I model my own voice in rvc gui or is there a simpler way?

[covert lake]

already replied in #✨│ai-help

[covert lake]

rvc gui is old, replied in #📑│making-models

[late nest]

Is there any way to integrate these models into https://sparkengine.ai ?

[lethal arch]

searsh megaphone efect pls broo

[late nest]

It's this no-code editor my team and I made. We want to integrate more technology through APIs

[lethal arch]

megaphone efect voice pls searsh

[final meteor]

Can any voice model pronounce the alphabet correctly in English? I tried and none can say the letter E and G for example
I also tried to say a singular word, "Me" for example and it fails at it

[onyx plover]

Is there a way to use the zip formats voice changer from voice to voice using the phone ? I remember it was possible but been a while since I didn’t use it

[final meteor]

Ok, and how do I get a voice model I downloaded and would like to use, to do that?
I have no AI knowledge

[chilly lake]

please re-state your question. What do you want to do?

[final meteor]

Well, if I try to say the same letter twice in a row, it will say the letter in such a awkward manner it's not even recognizable

[chilly lake]

are you talking about the realtime voice changer?

[final meteor]

Yes, rvvc 2.0.73--beta cuda

i can say abcdefg, but if i say aa or ee, it sounds really off

[chilly lake]

👉 #🔍│help-w-okada

[median kettle]

can someone help me for voice model create

[compact owl]

hey

[barren mauve]

hi

[compact owl]

where can i find the face swap collab

[barren mauve]

idk

[compact owl]

cmon man

[barren mauve]

hmm

idk wat u are talking about?

[compact owl]

you hurt my heart

[fallen basalt]

this is your monthly jay message

[white schooner]

stfu

[copper grail]

Why when my models sing high notes they break voice, no matter I train them using 50 epoch or 200 epoch?

[chilly lake]

if you use a proper pretrain that wont happen

[drowsy zephyr]

Sorry for the late response someone told me they have been using it but have no idea if this is true or not

[polar flax]

mostly a scam or like some "half life 3" unofficial mods

https://tenor.com/view/half-life-3-half-life-happy-meal-dancing-gif-18275228896923215828

[copper grail]

How do I do that?

[chilly lake]

for mainline rvc https://discord.com/channels/1159260121998827560/1265083834039533588

[copper grail]

Ty

[compact owl]

im tryna get my freak on

[clever crest]

Brody you good?

[jaunty fractal]

can somone how to dowload the voice changer\

i will give 10 dollar giftcard

on god fr

🙏

[polar flax]

go to #🔍│help-w-okada to check the pinned guide

[spark acorn]

hi

[ionic pumice]

damn that;s crazy

[spark acorn]

why

folks want to have fun

[ionic pumice]

you know damn well ts not what it meant

[spark acorn]

oh

[elder willow]

Guy Dm for fun now
DMs open for fun now
Dm for fun

[dawn temple]

What fun?

[spark acorn]

biology fun

[dawn temple]

[spark acorn]

It's an alternative account.
becareful !

[dawn temple]

Sure!

[spark acorn]

yeah
member since 12 feb

[dawn temple]

😮

[dawn axle]

Best image generator I can use rn for free? Need it for uni

[covert lake]

To generate images for free (text2img), either:

• Use @elder willow in https://discord.com/channels/1159260121998827560/1202754985255764060 (It's powered by DALLE3, from ChatGPT+), pretty easy
• Another easy and good ways with weighs.gg are:
  • Use /image with @hidden grotto in https://discord.com/channels/1159260121998827560/1202754985255764060
  • Create an image on their site https://www.weights.gg/ (which you can also use LoRAs, Low-Rank Adaptations, basically a small trained additional model to adjust your generation)
• Use Open Source Models like stable diffusion & flux that could be a bit **harder **but good, what's ur pc gpu? As you could run them locally (on ur pc) or on cloud (remote good pc)

[hidden grotto]

:wave: @covert lake, How can I help?

Available Commands:
• @weights find <query> or /find <query> - Search for RVC Voice Models
• /create - Create an AI Cover
• /image - Generate an Image

[hollow oasis]

i recently tried fooocus but my pc specs are so bad it couldnt cope

[covert lake]

what’s ur pc gpu and cpu

[hollow oasis]

no gpu

i got a laptop

so it cant really work

[median kettle]

If I record someone's voice, I can't see if anyone can model that sound. Just bring it to index and pth format

[covert lake]

it can

https://github.com/rupeshs/fastsdcpu

[covert lake]

u need to train it not just rename the format

[abstract sierra]

hi

[median kettle]

yes i know but i cant do this

i try maybe for 1 days

[torn hornet]

сосал?

[bold rampart]

Wdym, release this? You can download it via using a local AI apps installer, https://pinokio.computer

[tacit isle]

Late Ahh Response

[hollow spoke]

what ai voice changer software do you guys use

[covert lake]

Realtime for calls? Wokada, especially the deiteris fork, tell ur PC GPU *in #🔍│help-w-okada *

[ionic pumice]

lmfao

[hollow spoke]

Ok

[solar torrent]

You can do Stable Diffusion locally on a PC with no GPU but only CPU. Just don't expect it to generate fast.

[hollow oasis]

do we need to download models from stable diffusion

[solar torrent]

Stable Diffusion is not a website service where you can download a checkpoint model. It is an AI image generator program. You may need to download and put the model into folder separately.

CivitAI provides most of Stable Diffusion models you can download there.

[hollow oasis]

okayy i will try it

[tranquil kite]

Anyone knows where I can find those old models from Applio?

[covert lake]

wdym those old models from applio?

from their site before they took it down?

you can't, anyways iirc the majority were the same of the ones in #1175430844685484042

[tranquil kite]

It still missing what I want that used to be in Applio link

[covert lake]

You can search rvc ai voice models at:

• #1175430844685484042
• In #🔍│find-models , Do /find with @hidden grotto
• https://weights.gg/ (login required)
• https://huggingface.co/models (but watch out cus in hugging face there arent only rvc ai voice models)
• https://voice-models.com/
• https://thevoicemodels.com/ (for Turkish Models, login required with discord and level 2 on their server)

if there isnt one, you can:

• #1159289738314919936
• #1191429836321849435
• make it yourself with our docs guides https://docs.ai-hub.wtf/essentials/how-to-make-voice-models/

[hidden grotto]

:wave: @covert lake, How can I help?

Available Commands:
• @weights find <query> or /find <query> - Search for RVC Voice Models
• /create - Create an AI Cover
• /image - Generate an Image

[covert lake]

I'm guessing you were talking about the models in applio.org so I gave u all sources to find models since that site will never have a model page again

[tranquil kite]

But do you know why all of it got removed? And when?

[lime raven]

any of you know how long a 20 minute dataset usually takes to train

[brisk birch]

You know any website or something where I can use an a.i voice like I want to do stuff dennis.main is doing how do I do that

[lime raven]

i dont know who dennis.main is

but i think youre talking about weights.gg?

[brisk birch]

if u check on youtube I think you'll recognise the stuff he's made

he makes the goku skits etc

dap up tournaments

etc

[lime raven]

there it is

[brisk birch]

no but I dont believe

he uses parrotai

i checked it

it was GARBAGE

So I dont think its that

not the same quality at all

[lime raven]

sounds like he uses a tts kind of ai voice

maybe elevenlabs?

they have similiar intonations

[brisk birch]

I also thought that

but then theres a lot of parts

where they sing

and it fits hella well

😭

and the emotions to

when they're angry

or the way they talk

etc

[lime raven]

elevenlabs speech to speech?

i

if even not that then i dont know

[brisk birch]

text to speech

or

voice changer

wdym

[lime raven]

speech to speech is voice changer

like rvc is basically

[brisk birch]

oof

thanks

for ur help anyways

this shi hard to find

[lime raven]

ai is the future

#metaverse

[brisk birch]

@lime raven how do I use the voice models in this discord

ra

[amber hemlock]

So, Random quesiton. How many people here use suno?

[lime raven]

to use the voice models, you must download the rvc gui

[brisk birch]

oh ok and thats it?

[lime raven]

and watch a tutorial on how to use it

[lime raven]

import the voice model into it, yada yada

or alternatively

you can go to weights.gg which has voice models too

that you can use without having to install any 3rd party stuff

its all in-website

but you wont get 48k quality models on there

[brisk birch]

ok so so

if I download rvc gui

I can use the voice models here

and I dont have to train anything

etc

[daring saddle]

I have been having issues with the AI Voice shiz

Omg I am so blind

Theres a help channel

Right in front of me

[brisk birch]

why is it whenever a cuss word pops up on the ai it smothers the word lol

#✨│ai-help

[supple horizon]

Hey all!

I have audio files of a single person (monologue basically) that is in german and I would like to voice clone it into italian.

Is there a good guide here somewhere I could follow to achieve this?

[gaunt light]

hi im trying to find the most realistic rvc voice models but ones i find online is not always the best quality. Can someone help me?

[balmy fern]

em what sigma

[minor blade]

Check the #1175430844685484042 channel and test which one fits you better.

[glass dock]

hey, does anyone know how to make those ai videos of celebrities talking over a script?

like lebron talking over a video about how some math works

[polar flax]

the old one from tiger18n repo?

[elder willow]

DMs open for fun now
Dm for fun
HMU for fun now

[old haven]

anyone know any ai apps?

for voice changers

[elder willow]

Guy Dm for fun now

DMs open for fun now

[solar torrent]

I don't take a direct message from a random person I don't know.

[ruby wharf]

hel no

no one dming you pal

[solar torrent]

Just letting you know, I don't Diddy people.

[ruby wharf]

me neither glad we on the same page

imagine diddling in the big 2025

could not be me

[pallid socket]

anyone here

I need help

[jaunty fractal]

yo

slide some

i dont know how to dowlaod

it needs tutorial video

[covert lake]

there’s no youtube tut, also I replied to u in #🔍│help-w-okada and RVC gui is old

@brisk birch rvc gui is old, tell what u are looking to do and what’s ur pc gpu and cpu in #🔍│help-w-okada

[covert lake]

!howtoask

[pine acorn]

How To Troubleshoot

__**GIVE CONTEXT.**__ 📝

• Don't simply mention your issue, like "my rvc is not working".
• Describe the step you are on, what you're trying to do, the RVC you're using, a screenshot, etc.
• The more context, the better.

__**BE POLITE.**__ <:matsuripray:1159685390156967936>

• Don't be desperate. You can ping a Helper, but if they ignore, they aren't available/don't know the answer.
• It's okay if you're frustrated, but don't take it into this server.
• Don't DM without prior consent.

__**BE PRODUCTIVE.**__ 🤝

• Don't ask for every little instruction. Put your own effort & test things by yourself.
• Don't ask to ask.
• Check if your answer is a Google search away/on our guides website.

If you've changed your mind about your message, feel free to delete/edit it.

[covert lake]

also use the appropiate help channels and elaborate pls

[covert lake]

realtime for calls? Tell ur pc gpu and cpu in #🔍│help-w-okada

[covert lake]

You can search rvc ai voice models at:

• #1175430844685484042
• In #🔍│find-models , Do /find with @hidden grotto
• https://weights.gg/ (login required)
• https://huggingface.co/models (but watch out cus in hugging face there arent only rvc ai voice models)
• https://voice-models.com/
• https://thevoicemodels.com/ (for Turkish Models, login required with discord and level 2 on their server)

if there isnt one, you can:

• #1159289738314919936
• #1191429836321849435
• make it yourself with our docs guides https://docs.ai-hub.wtf/essentials/how-to-make-voice-models/

[hidden grotto]

:wave: @covert lake, How can I help?

Available Commands:
• @weights find <query> or /find <query> - Search for RVC Voice Models
• /create - Create an AI Cover
• /image - Generate an Image

[tough moss]

how fast do RVC models work?

[covert lake]

it depends on ur pc gpu

.what is it and what are u looking to do

[tough moss]

translate sounds to make them more human

I’ve got 2 NVIDIA Quadro 5000s

[covert lake]

translate sounds ?

I don’t understand

[tough moss]

aren’t RVC speech to speech?

[covert lake]

rvc is speech to speech, but wdym sounds

like u just want to take a random sound and make it a human?

[tough moss]

no

should’ve worded it better

I meant speech generated from a TTS model

then through an RVC

[covert lake]

Ohh

yeah u can do that

u got rvc already or need help for installing it

[tough moss]

Just wondering if it will increase time taken by a lot

[covert lake]

Well not much, for example in Applio they have "built in" tts, which makes an audio tts via Ms edge API then uses it as an input

[tough moss]

I see. On the topic of TTS, which models are a good balance between realness and speed?

[polar flax]

might be useful for training, but even a GTX 1060 or a cpu can still do inference

also you should mention the quadro generation (whether pascal, turing, or ada) since it could be confusing

[covert lake]

like which one of the tts models in applio ? Or talking about any tts?

[tough moss]

any in general

[covert lake]

The fastest one is PiperTTS but it isn’t really as good in quality

[tough moss]

Would PiperTTS through RVC yield good results?

[covert lake]

maybe you could try Fish Speech ot F5 which are 0 shot (no training required, unliek rvc which is few shots and need training)

[covert lake]

eh haven’t tested that

if u want there’es https://docs.aihub.gg/tts/tts-tools/

[tough moss]

I’ll have a read

Thanks

[covert lake]

I would suggest pipertts only if you desperately need speed, but it’s worse in quality

also yw

[tough moss]

I’ll see how PiperTTS works with RVC

does PiperTTS have a hugging face link?

[covert lake]

https://github.com/rhasspy/piper

[tough moss]

thanks

[brisk birch]

u know anything that give better results than rvc gui cause some times the words dont match right or sound muffled and I keep seeing on youtube how these ai voices sound hella good

[covert lake]

rvc gui from t4ger on github?

that one is old

tell ur pc gpu and cpu and what u are looking for

[brisk birch]

oh

how do I check that

[covert lake]

You can check your pc gpu via:
ctrl+shift+esc (task manager) -> Performance tab -> GPU

[brisk birch]

Im looking for good enough quality where I can make shi like Timonsdreams or dennis.main

[covert lake]

RVC = Retrieval-based-Voice-Conversion, the best Speech To Speech AI Models (on v2), Inferences (use models) pre-reocrded audio (ai covers) and train (make) models

Wokada = uses RVC for realtime inference

which one do u want

[brisk birch]

AMD RYZEN 7 5800 With radeon graphics (CPU)

RADEON RX 5500M GPU 0

AMD RADEON (TM) GRAPHICS GPU 1

[brisk birch]

whatever is good enough where I can make the content like dennis.main or timonsdreams 😭

[covert lake]

idk who those are

[brisk birch]

fr?

crazy

[covert lake]

read that message, and say which fills your needs

[brisk birch]

https://tenor.com/view/dennis-main-dbz-gta-barber-perfect-dap-universal-dap-gif-895684480432045778

this guy

oh

[covert lake]

i explained you what they do

they are different rpgorams

they both use the same model architecture

[brisk birch]

which one gives the best voice for the character like I put a mp3 file and it says it the best or sing it etc

[covert lake]

then rvc which does inference (use models) on pre-recorded audios

[covert lake]

ehh

Your AMD GPU is good enough to do inference (use models) locally (on ur pc), you won't be able to train (make models) but use them

You can:

• Locally (runs on your pc so the speed depends on that, you will have to set it up with the guides):
  • Applio (AMD Windows) : A fork of RVC with some extra features like Applio TTS, kinda faster and simpler but same quality tho
  • Mainline (AMD Linux/Windows) : The original RVC
• Cloud (remote good pc, easier and faster than ur PC but it's limited):
  • Ilaria RVC Zero: fastest and simplest that you can get for free
  • Weights.gg: Partnered with AI Hub, lets u do them easily but u may be in a queue
  • Applio Colab: max 4 hours daily, not granted, of GPU

Easiest possible (automatically separates vocals & instrumentals) : weights.gg
easiest cloud: Ilaria rvc zero
easiest local: Applio

[brisk birch]

ok how do I get this

[covert lake]

I would just suggest using cloud

I gave you all links

[polar flax]

most videos about rvc are too old and it might have some editing magic that may sound too good unlike the reality

[covert lake]

following the cloud links

[brisk birch]

u know what they use to boost it then?

[brisk birch]

thanks unc

[covert lake]

How old are u

[brisk birch]

18

[covert lake]

I’m literally 16 😭

[brisk birch]

damn my fault youngin 😭

[tight cypress]

got the GPT SoVITS v3 hit and running

v3 doesn't like torch.float16 for some reason

btw v3 switch to DiT compared to VITS in v2 and v1

[chilly lake]

where did you find v3?

[tight cypress]

https://huggingface.co/lj1995/GPT-SoVITS/tree/main

the creator updated the repository

[elder willow]

Guy Dm for fun now

Add me as a friend for fun now

DMs open

[ionic pumice]

damn that's crazy

[ionic forge]

What's the best voice for narration?

I want to create a voice message for my student as English listening test.

Plz be American accent

[brisk birch]

u guys know where I can find RVC Files that I can download

index etc

[wild chasm]

Is there a Realtime Voice Changer for free?

[ionic pumice]

-rt

[red frost]

Interaction has expired, use the command again for a new interaction.

💻 Local Realtime RVC

• Guide for deiteris' modified W-Okada's Voice Changer Client (Fork)
• Guide for W-Okada's RealTimeVoiceChangerClient
• How to use real time conversion with RVC (voice changer)
• RVC TTS

[covert lake]

Tell your PC GPU and CPU in #🔍│help-w-okada

[covert lake]

I think you mean models

You can search rvc ai voice models at:

• #1175430844685484042
• In #🔍│find-models , Do /find with @hidden grotto
• https://weights.gg/ (login required)
• https://huggingface.co/models (but watch out cus in hugging face there arent only rvc ai voice models)
• https://voice-models.com/
• https://thevoicemodels.com/ (for Turkish Models, login required with discord and level 2 on their server)

if there isnt one, you can:

• #1159289738314919936
• #1191429836321849435
• make it yourself with our docs guides https://docs.ai-hub.wtf/essentials/how-to-make-voice-models/

[hidden grotto]

:wave: @covert lake, How can I help?

Available Commands:
• @weights find <query> or /find <query> - Search for RVC Voice Models
• /create - Create an AI Cover
• /image - Generate an Image

[covert lake]

GPT so vits V3 before of RVC v3 lol

[chilly lake]

i did check... what a mess

stealing shit from other projects 🙂

bigvgan

and DiT from F5-tts

[polar flax]

iirc bigvgan, bigvsan, and evagan seemed to be sota, and then the bigvgan v2 was out in the middle of last year

[chilly lake]

BigVGAN 44k 512 upsampling ratio....

44100/512 = 86.132

does not compute

I think the main difference is the cuda kernel

new mel loss (already have that in Applio), cqt discriminator - does shit all

[covert lake]

Wait What?

[chilly lake]

"stealing" maybe a wrong word, but..

dont think the diffusion transformer is useful for us

old VITS is what is used to turn a spectrogram into latents and latents into a spectrogram

[glad nebula]

so thats what rvc boss was into all of this time of not saying anything

killing rvc for that 😭

[chilly lake]

well, you see... you done one project for university graduation, then to go to phd and do a better one

[glad nebula]

i mean yes but hurts seeing rvc getting killed just like that

im aware he did that in order to do something better

the original rvc was a mess

i see why he doesn't want to be involded in that anymore

[quartz roost]

Well diffsinger is getting new development

[minor blade]

I never liked diffsinger, utau neither vocaloid 😄

[quartz roost]

Why?

[minor blade]

🐢 👍

[ionic pumice]

https://media.discordapp.net/stickers/1277491531195875421.png?size=160&name=thats+kinda+crazy+gang

[quartz roost]

Well it's kind of meant for a different purpose

[minor blade]

👍

https://www.youtube.com/watch?v=egcVBp_5XnQ

@calm vapor Listen to this

[covert lake]

ah f5

[covert lake]

yup

[covert lake]

I mean I think that's the wrong word

[chilly lake]

borrowed

without sking for permission 🙂

[misty ferry]

Hi guys, anyone here has heard about kluster.ai? I've been wanting to give it a try but wanted some opinions first

Sorry if asking this is against the rules, I'm new here

[solar torrent]

I've never heard of this website name, sorry.

[covert lake]

From what I understand, it's using F5 code right

[trim tinsel]

Mcdingle fartling

[solar torrent]

One of the only websites I remember I did some things is Weights.gg.

[trim tinsel]

Ur new name is Mcdingle fartling

[solar torrent]

But I'm not Quandale Dingle.

[chilly lake]

the diffusion transformer

[ionic pumice]

vocaloid1-quality song

[solar torrent]

There doesn't seem to be any place here to promote your things now.

[charred ermine]

thai

[covert lake]

Oh 😭

Well technically F5 tts is MIT licensed, tho this is really weird

I don't understand why basically every AI project is MIT licensed tbh

@chilly lake https://f5tts.org/playground do you know about this site

I can't find anything about it in the F5 official github

I feel like it's fake, the playground is just embedding the hf space demo

[chilly lake]

probably

There's new contender in tts called Zonos

took almost 3 hours to instal it on windows

seems interesting, does voice cloning and emotions

[quartz roost]

So I'm still waiting in an endless queue just trying to hear a sample

Where is the local version

[covert lake]

https://github.com/SWivid/F5-TTS

[covert lake]

How's it?

Also did u hear about kokkoro?

[covert lake]

Damn, I thought u wanted to switch to linux

[quartz roost]

Thanks, I'll look when I'm at my pc. No gui?

[covert lake]

There's a web UI and cli mode

[quartz roost]

Thanks, I'll experiment later

Ah it uses whisper too

[chilly lake]

both kokoro and zonos use espeak as phonemizer, kokoro has no voice cloning, but there are some voices included.. not much emotion, but pretty good reads. zonos has 30s segment limit (~500 characters), otherwise it starts hallucinate, voice cloning is decent, emotions too

[covert lake]

Do u think it's worth adding them to the docs?

Also thanks for the clarification

[chilly lake]

kokoro definitely yes

zonos i'd wait until there's easier windows install that only requires torch and no other bs

[quartz roost]

https://www.zyphra.com/post/beta-release-of-zonos-v0-1

Has anyone tested this? I saw mention or pricing, is there a local version to render for free

[chilly lake]

read above

[quartz roost]

Oh, I see

[chilly lake]

right now you need python 3.11, cuda torch, triton, flash attention, mamba_ssm

last 3 require some effort to acquire on windows

[covert lake]

Mm I will see

[quartz roost]

I like the inflection of the zonos demos

[covert lake]

I was also thinking of adding Piper and explicitly say it's only for people who want speed, since I saw some asking for fast ones

[calm vapor]

this fire

damn

[minor blade]

It was made by my fellow friends. 👀

I made the models for them hehe.

[calm vapor]

I need to hear standly ai bro

i got a 11/11 model but its mid

not mid

but its okay

you ythink that the no te valoro guy can cook with it?

[urban bridge]

Si vous devez revenir à 14 ans, qu'est ce que vous feriez différemment pour changer votre vie actuelle (choix, projet, couple, amis etc...) ?

[chilly lake]

how fast is piper?

[covert lake]

Fast enough to even run fine and optimized on a raspberry pi 4

[chilly lake]

so basically pure espeak?

kokoro takes <4s for 1600 symbol text

beats f5 for sure

[covert lake]

they use their own phonomizer https://github.com/rhasspy/piper-phonemize the models are trained with VITS and converted to onnx

[chilly lake]

"When using eSpeak phonemes, requires an espeak-ng fork"

[covert lake]

yeah it's a fork

tho, it's more lightweight and faster since it's made by an open smart homes foundation

[chilly lake]

tried Zonos with long text, very bad

[covert lake]

[chilly lake]

probably need to set the seed before every sentence or something

too cheerful

[covert lake]

looks kinda weird at the end

[chilly lake]

since it does not support long sentences I had to split and then stitch them together, but it sounds really weird and disconnected

[covert lake]

agreed

[olive rampart]

GOYDA

[chilly lake]

xtts is the best story teller

[olive rampart]

goydaaaaaaaaa

[lime raven]

are pretrains basically like mannequins and the datasets we provide them are like clothes?

[minor blade]

Nah bro i don't think so

He's currently signed to a label called Moneywayy

NG is currently occupied with his irl job and making AI demos for a producer from the label.

(I don't wanna be rude but we also don't care too much about Standly)

[chilly lake]

pretrains are like half-made marble statue where you just finish the finer details. It is very hard to reshape a finished statue to another shape

[lime raven]

i see

[chilly lake]

well, even more, it is a collection of half-finished statues of varios shapes

and you pick one to finish (singer) and break the others

[lime raven]

hm

okay

[onyx oxide]

can someone point me to the download for windows i look at the forum and i cant find shit idk if im dumb

[chilly lake]

download what? realtime voice changer?

[minor blade]

!howtoask

[pine acorn]

How To Troubleshoot

__**GIVE CONTEXT.**__ 📝

• Don't simply mention your issue, like "my rvc is not working".
• Describe the step you are on, what you're trying to do, the RVC you're using, a screenshot, etc.
• The more context, the better.

__**BE POLITE.**__ <:matsuripray:1159685390156967936>

• Don't be desperate. You can ping a Helper, but if they ignore, they aren't available/don't know the answer.
• It's okay if you're frustrated, but don't take it into this server.
• Don't DM without prior consent.

__**BE PRODUCTIVE.**__ 🤝

• Don't ask for every little instruction. Put your own effort & test things by yourself.
• Don't ask to ask.
• Check if your answer is a Google search away/on our guides website.

If you've changed your mind about your message, feel free to delete/edit it.

[onyx oxide]

alright so i got the file but how to i unzip it like ive unzipped in winrar but nun comes up

[frigid lantern]

For those who use W Okada, is client better than server?

[kindred kelp]

My legs are getting stronger. I'm paralelgic from a gunshot wound but I'm getting better it's been 4 years

[plain cove]

https://cdn.discordapp.com/attachments/1233711590965121086/1335289286043369482/togif_2.gif

local

[elder willow]

DMs open for fun
Add me as a friend for fun now
HMU for fun now

Guy Dm for fun now

[barren mauve]

this is not a hook up server

[polar flax]

https://cdn.discordapp.com/attachments/1159289516268474570/1339753909786378310/image.png?ex=67afde8f&is=67ae8d0f&hm=038ddb00f404b57c6a5e20fd05c28a1105be2872434646fe9315210de534c9f6&

[barren mauve]

i know right

[tepid basin]

Does anyone know a good instrumental generator? @barren mauve is looking for something to generate a film score. Something a little better than Suno if possible

[solar torrent]

Skibidi Toilet. 🔥

[solar torrent]

Imagine being so lonely online, you gotta expect someone to talk to you.

[polar flax]

seems more likely a bot

[solar torrent]

When I use "Arona President" voice model instead of a voice model of Arona, it sounded like some girl who tryna sound like Arona.

https://cdn.discordapp.com/emojis/1015884939406868561.webp?size=48

[gilded fractal]

@graceful patio why is bro in this server as well 😭 🙏

[solar torrent]

Is she interested using voice changer to troll some "kid" after all? I'm pretty sure she is.

[ionic pumice]

what

where the hell did you even get that assumption

[polar flax]

https://tenor.com/view/huh-cat-gif-26460616

[solar torrent]

There have been some dramas against her recently, especially when she was caught leaking some kid's phone number last year.

[sage coral]

okay how does this work haha

[covert lake]

how does what work?

what do u want to do

what’s ur pc gpu

There’s thousands of ai programs and models

[covert lake]

Ragebait

[versed jetty]

hey any one to talk with me

iam arthur morgan

this is a big chance to talk with me bastards

dutch tell me about something named as chatting idk

ok fu!@#$ you bastards go away i dont love chatting any way

[polar flax]

he prob means that ragebait drama above

[frank shell]

hello, which collab separates and joins automatically?

[polar flax]

im not sure what you exactly mean but that sounds creepy

[frank shell]

for example, I put the voice model in collab and then collab separates the audio from the music and after doing the work it puts it all back together again

[tepid basin]

https://github.com/audacity/audacity/releases

[river verge]

DIlly ding, dilly dong! A new RegalHyperus drum model just released!
Supernova (Drum model no. 579)

[quick pewter]

Someone know why RVC V2 on colab is not working?

i was trying to train a model

[random palm]

wrong chat mb

[fossil rover]

это что

что за чат

#🧬│ai-chat

ак

ак

[pliant elm]

пожалуйста, говорите здесь только по-английски, в противном случае используйте #🌏│русский

[covert lake]

In #✨│ai-help tell:

• ur PC GPU
• what google notebook u are using (link)

[weak blaze]

Hello

Question, is there a way to replace my android tts voice with a voice model from here, if so how?

I dont know to ask in help considering how small of a question this is

[waxen tartan]

is there an AI to improve the quality of vocals?

[kindred kelp]

Jeez why does everyone find a.i. vocals and music so interesting. We really should be focusing on the a.i. code that's possible not pretending to be a girl on Omegle or making a new a.i. drake release

[grim flax]

selam

[tough moss]

Currently using LM Studio and thinking of changing to AnythingLLM, Ollama or Open WebUI
any opinions or experiences?

[covert lake]

Open WebUI is only the Web User Interface, it can't work alone, it needs to be in pair with another tool like Ollama

I personally prefer Ollama, easy to use and Open Source

[tough moss]

Oh I see. By the way, does Ollama perform better than AnythingLLM or LM Studio? Or is it just because it is open source

does Ollama have its own UI?

[covert lake]

I haven't tested AnythingLLM or LM Studio, but I have heard from others that LM Studio might be slower https://www.reddit.com/r/LocalLLaMA/comments/1c18hgj/why_ollama_faster_than_lmstudio/
I can't confirm that tho, but I care about transparency

[covert lake]

ollama is cli and also can be used as an api, you can just use Open WebUI which is originally meant also for ollama

[tough moss]

I see. I’ll keep that in mind and do my own tests.

I’ll have a look at Open WebUI & Ollama together

[covert lake]

Great, let me know about your tests, I'm interested

[rapid canopy]

hello

[wraith hull]

rus discrimination

[ionic pumice]

https://media.discordapp.net/stickers/1277491531195875421.png?size=160&name=thats+kinda+crazy+gang

[pliant elm]

bruh

[ivory cove]

hey anyone know how to set up an ai to read for me ?

[digital peak]

just use immersive reader

i don't know really, haven't been here for a bit

[covert lake]

isn't that what immersive reader does? or do you mean smt else?

[covert lake]

maybe he meant Text To Speech for content creation but I'm not sure

[ivory cove]

i want it to have a specific voice from a show i like, if thats possible

[covert lake]

okay so you want to touch a specific part of your phone to read it out loud with a custom voice ?

[ivory cove]

Mostly like online book pdfs

im fine for windows

[covert lake]

basically you want to make an ebook?

first you would need to use an OCR (Optical Character Recognition), like https://www.ilovepdf.com/ocr-pdf or https://github.com/tesseract-ocr/tesseract

this basically gives you the text from the pdf

then you would need a TTS tool

There are different Text To Speech (TTS) AIs:

GPT So Vits: RVC isn't as good as GPT So Vits for tts, but gpt so vits (few shot tts, which means needs just a lil training for models) can't use rvc models (and viceversa), and its only limited to: english, chinese & japanese, if you wanna check gpt so vits instead, read https://docs.ai-hub.wtf/tts/gpt-sovits/

Freemium 11labs: An easy way to do TTS is https://elevenlabs.io/, you can't use RVC model on this but its a mostly premium easy way for good quality TTS

FishSpeech: FishSpeech is a 0 shot (no explicit training needed) TTS, if you got a good pc you can use it locally else use their site

With RVC Models:

RVC is natively for Speech To Speech, but forks such as ilaria rvc mainline & applio have built in tts (using Microsoft Edge TTS to make a generated tts audio, which i suggest you to choose a tts model that is the same gender and language of the rvc model you wanna use, and then convert it with rvc)

If you wanna do tts locally with RVC Voice Models (if you got a good pc):

• You can get Applio in our docs
• While Ilaria RVC Mainline here (no guide as of right now)

If you don't got a good pc you can do tts with RVC Voice Models on cloud:

• Ilaria RVC Zero (Running on A100 GPU, free fasted rvc on cloud) and the guide

• Use Applio UI Colab (with google colab T4 free daily limit gpu)

• if you don't wanna use edge tts, you could try another tts ai from our tts index and use the output as an input in rvc

I have never did this myself for making an ebook but this info could help you out

[astral island]

hi

[covert lake]

hello

[elder willow]

Guys, how do I find more GPT Sovits Model?

[gentle crag]

I am using deepseek in ollama in anything llm ,so if deepseek gets new version, how do i update the deepseek model

[heady osprey]

hey, how do i upload my ai model

to the forum

[clever ridge]

p^m$*

[barren mauve]

hello

[covert lake]

you can check #1175430844685484042

[covert lake]

to #1175430844685484042 ? check https://docs.aihub.gg/extra/model-maker-role/

[heady osprey]

ok

[covert lake]

it's not like a program (like discord) that gets auto updates, if deepseek releases a new model, you gotta download the new model

[timber river]

ok

[minor blade]

You're not the real bella the wolf, right?

💀

[hollow tangle]

yo you single hmu

[dapper ginkgo]

💀

[barren mauve]

she might be a minor

[dapper ginkgo]

She isnt LOL

[hollow tangle]

shes a youtuber shes not afaik

[barren mauve]

ok

[hollow tangle]

im insane but i have standards

[barren mauve]

yeah lol

im still single tho : D

[hollow tangle]

[dapper ginkgo]

[barren mauve]

XD

[covert lake]

u sure

[hollow tangle]

THE FUCK??

[barren mauve]

XD

[hollow tangle]

source??

[covert lake]

https://budsforbuddies.fandom.com/wiki/BellatheWolf

not sure tho

anyways I doubt that's even her official acc

[barren mauve]

hmm

[covert lake]

it's just a tiktok troll anyways

[hollow tangle]

official links are linked tho

and she surely doesnt seem 14 from her recent video apology

if shes really a minor damn my bad

[barren mauve]

im adult tho

[hollow tangle]

what should i do with this information

[covert lake]

damn

[barren mauve]

idk

[covert lake]

wait, she's a minor and posting nsfw on twitter?

[hollow tangle]

what

[barren mauve]

yep

[hollow tangle]

😔🚬

[covert lake]

atp I feel like this is just a guy trolling for some little reactions lol

the acc died too since years anyways

[barren mauve]

maybe

[hollow tangle]

https://tenor.com/view/colleen-ballinger-miranda-sings-youtube-unban-youtube-drama-gif-16045623481857669326

[barren mauve]

u make models?

[hollow tangle]

yea

[barren mauve]

noice would u like to help me out?

: D

[hollow tangle]

sure

[barren mauve]

yay ^^

[hollow tangle]

how can i help you?

[barren mauve]

idk how to create one i been trying but im stuck

😦

[hollow tangle]

follow guides

[barren mauve]

i tried it

im just confused lol

[hollow tangle]

by?

[barren mauve]

how to make a voice model

[hollow tangle]

gather a dataset first

[barren mauve]

how?

were can i get the database?

i meant dadaset

[hollow tangle]

wdym how, it really depends on what voice you wanna make

[barren mauve]

dataset

a deep voice similar to corpse husband

: D

[hollow tangle]

find it on youtube

download it

clean it manually or with uvr

train it

[barren mauve]

ok

how?

idk how to train a model tho

:/

[dapper ginkgo]

what the holy fuck

[minor blade]

💀

[covert lake]

damn

skill issue tbh

[vagrant sparrow]

Hi all, I want to change my voice in real time. I have an RX 6600, will it be enough for this, and what app do I need for this?

[minor blade]

You can start by using Deiteris' w-okada fork.

Lemme give you the docs.

https://rentry.co/ForkVoiceChangerGuide

For further questions go to the #🔍│help-w-okada channel

[vagrant sparrow]

Thanks

[minor blade]

You're welcome.

[night lake]

So like... Who are you? And why are you here

I feel like I've heard some horrible drama but IDK if it's real or what the drama was

[severe heath]

Привет

[gilded fractal]

Nothing just wondering lol unban me diddy

https://tenor.com/view/eye-of-rah-gif-10095752177411352905

[polar flax]

@icy pendant correction in the doc: the valentine's day has passed and still no confirmed torch fix for RTX 50-series. perhaps until the RTX 5070 Ti or RX 9070 XT is released

[static inlet]

can someone help me with the voice changer stuff

[glass junco]

nooo way less goooo

[fiery kettle]

are u really taking the gemini summary at face value

[glass junco]

i mean open ai did say that its going to kill there product i dunno

[solar torrent]

Now read the reference website blog of it and then summarize.

[worldly lava]

where the download for the voice changer

[verbal shale]

Do anyone have udio ai (I'm making a youtube song but its family friendly)

plz

[solar torrent]

This channel here is not where you asking for W-Okada. For W-Okada, go to #🔍│help-w-okada

[solar torrent]

No, I'm not giving my account to anyone. Search Udio on Google.

[molten crag]

Hi

[fervent fjord]

Hlo i want to know why is this server created i mean what are the benifits i may get in this server

[ionic pumice]

none

[solar torrent]

AI cover, but now it's mostly AI voice changer.

[elder willow]

hi

[knotty meteor]

wassup, can someone train a model for me? I already got the dataset

is there a specific channel on here to ask for that?

[solar torrent]

To request someone to do a voice model for you, you can do one in #1159289738314919936

[knotty meteor]

thanks

[ionic pumice]

namari would you be the ts to my pmo

[jaunty fractal]

Give me the voice changer

Pls

[polar flax]

https://tenor.com/view/nuh-uh-beocord-no-lol-gif-24435520

[solar torrent]

What does this mean?

[solar torrent]

This channel isn't where you asking for W-Okada. I'm not Santa Claus.

[wraith hull]

what a hottie replied to me

[icy pendant]

Deiteris said he could make a fix happen, thats all

[ionic pumice]

IM CTFU

[polar flax]

btw someone said that torchaudio is the culprit; it doesn't support the latest nightly torch, hence falling back to the stable torch 2.6/older which the RTX 50-series cards refuse to run

[urban bridge]

What free formation or videos did you recommand me if i want to start an AAA agency ? (i am a beginner)

[keen marsh]

can someone explain to me what pretrained models are?

[covert lake]

they are basically a base for training your own models

[keen marsh]

is it better than just using your own model? also how can i train a model from vocals? i can't find anything in #1159513888199540817 maybe i'm blind

[covert lake]

every model uses a pretrain, you wouldn't train without one bc it would take a massive amount of data and time

what's ur pc gpu

[keen marsh]

1660 super

i can't pretrain through a website?

also what is the best pretraining model to use?

[covert lake]

Your Nvidia GPU is good enough to do inference (use models) locally (on ur pc), not the best to train (make models) even if still possible

You can:

• Locally (runs on your pc so the speed depends on that, you will have to set it up with the guides):
  • Applio: A fork of RVC with some extra features like Applio TTS, kinda faster and simpler but same quality tho
  • Mainline: The original RVC
• Cloud (remote good pc, easier and faster than ur PC but it's limited):
  • Ilaria RVC Zero: fastest and simplest that you can get for free
  • Weights.gg: Partnered with AI Hub, lets u do them easily but u may be in a queue
  • Applio Colab: max 4 hours daily, not granted, of GPU

Easiest possible (automatically separates vocals & instrumentals) : weights.gg
easiest cloud: Ilaria rvc zero
easiest local: Applio

[covert lake]

there isn't one, it depends by your model dataset lenght and language

[keen marsh]

my gpu isn't better than running it on the cloud no?

[covert lake]

your gpu is old, it can technically do both inference and cloud but it isn't better than cloud for training

I would suggest you to use cloud for training

[clear dock]

Hello

[regal drum]

any good alternatives for Suno ?

iknow a few of them but i was wondering if there is something new i dont know about yet

[covert lake]

udio

Weights.gg

[verbal shale]

I hate the new update of suno

[summer bluff]

Hello

[nimble arch]

What did they change? Idk, since it's been a while since I used it...

[covert lake]

hello

[summer bluff]

Are you ai or real

[covert lake]

As a Large Language Model, I dunno gng 💔

[summer bluff]

What’s 2+2= AI

[covert lake]

Hello mr Resalesmelo, it equals to 4.

[summer bluff]

Okay Ai I have couple question for you how many letters are in alphabet

[covert lake]

There are 26 letters in the Modern English Alphabet.

[summer bluff]

Okay what is my birthday ai

[covert lake]

Your bday is the 32 of February 2077

[summer bluff]

Wrong ai

[covert lake]

[agile condor]

you are an ai assistant by Deepseek. how many 'r's in "pretrained recurrent generative imitator"

[summer bluff]

So now I question you are you real or fake

[covert lake]

<think>
I'm not actually an AI, but the user wants me to be an AI assistant by Deepseek, so I gotta do it
</think>

There are 7 'r's in "pretrained recurrent generative imitator"

[covert lake]

Are u srs 😭

@tepid basin I should get a job acting as an AI atp, I think I'm pretty beliavable

[tepid basin]

Full time AI

Not AI dev

Just AI

[covert lake]

You're right, United shoes!

[polar flax]

prompt engineer, etc

[tepid basin]

@covert lake you and @opal marsh should fight

[covert lake]

Smhh I don't sound like an ai that much

Right?

[night lake]

you kindaaaa do

prime example: #💬│staff-chat message

[hard panther]

Hi, Im studying computer science with AI want to work in the AI industry (UK). What are some warnings or key considerations I should know about ahead of time?

[solar torrent]

If I were a fake ass person, I wouldn't have been here telling people how to install RVC/W-Okada, you know.

[verbal shale]

You can't add any bad stuff in the songs like funk or beach or sheet

[nimble arch]

Swearing? ON MY CHRISTIAN MINECRAFT SERVER?!

[wraith moss]

Correct me if I’m wrong, but AMD’s ROCm is pretty much ready to ship in windows… I was looking into it and the HIP SDK seems to have all the required calls and functions. Is a version of PyTorch built with the SDK the only thing missing, or am I a dummy and missing something else?

[covert lake]

That sounds like a great human message

[pliant elm]

last time I checked ROCm was silly in terms of what it supported, did that change?

[dawn temple]

Nick, why don't we ping you when we need help, you are a bot already 😂 /jk

[covert lake]

real

[dawn temple]

ahahaha

[chilly lake]

rocm works fine in Linux

or WSL2, which is linux

AMD's Windows support for AI is pitiful

but you can use HIP SDK + Zluda emulator to run most CUDA Pytorch applications

[calm venture]

hello guys, does it exist some ai that create for you an ai voice model based on some sample? Because i'm using applio but it would be cool to have a program that automatically create an AI voice model based on some sample o more and to be use after on Applio, thanks in advance!

[covert lake]

Replied in #📑│making-models

[solar torrent]

No.

[long spoke]

anyone knows what's the python code to put the korean-hubert-base as embedding model?

[wind narwhal]

Guys, who has a subscription to leonardo.ai?

[dapper trail]

Not me, but if I had to recommend a good image generator, I would recommend Foocus

[polar flax]

bro left

[dapper trail]

[sterile pecan]

Is there an AI song generator where I can upload an existing song (let's say a spanish song), then I'll provide lyrics in different language (english) and the AI will "sing" the song in english?

[dapper trail]

with suno you can change the words of a song, but I don't know if it will come out well, I do it sometimes but with songs in my language and I make them say other things

[sterile pecan]

Just checked it out, seems like its a premium feature?

[dapper trail]

no, it's free, but the song must be 1 minute long at most, otherwise you need the premium version

[sterile pecan]

I see, thanks! Hopefully, in the future, these things will be more accessible and more advanced. Only time can tell...

[dapper trail]

[surreal acorn]

Anyone have experience with CNNs?

[thin spindle]

Looking for a model maker to create a model asap. Will pay. Please send dm of your work samples!

[chilly lake]

RVC has a bunch of CNNs

[covert lake]

#1159289738314919936 or #1191429836321849435

[kindred kelp]

Why is this channel always flooded with people looking for voice mod? You guys should just pin the way to do it to the topic

I use play.ht

[covert lake]

We literally have a full documentation on it, and guides linked all in pinned messages 😭

[covert lake]

smh not open source

[kindred kelp]

So why are ppl still begging here lol

It's annoying. I come here daily and see the same conversation over and over

[night lake]

people refuse to read

[kindred kelp]

Can someone help me mod my voice to make fart noises every time I talk??

Jk

[stiff crest]

rah

[minor blade]

what

[kindred kelp]

I'm joking

[polar flax]

ppl used to ask before reading the rules and guides

[surreal acorn]

I'm looking for help developing an CNN for interpolation between frames for a Vid2Vid model. Just looking to chat with someone with some expierence

[polar flax]

I thought it's using transformers already?

[chilly lake]

only HuBERT model from transformers

[wraith hull]

The guy downstairs has so many dicks in his ass
↓

[tranquil lantern]

https://media.discordapp.net/attachments/983399166225047562/991456564282142891/dumbass.gif

[river verge]

DIlly ding, dilly dong! A new RegalHyperus drum model just released!
Sing Me to Sleep (Drum model no. 580)

[toxic crystal]

hello
have a nice day

[lilac ridge]

hlo guys

hrya

[covert lake]

hello, have a nuce day too

[covert lake]

hello

[covert lake]

i’m good wbu?

[cobalt coyote]

hmm

[lost crater]

guys, make pls "папич"

[nocturne bridge]

how to use intel arc gpu in voice changer?

[covert lake]

replied in #🔍│help-w-okada

[covert lake]

we allow only english here except in some language channels

[lost crater]

😦

[elder willow]

lets take a stab... 👀

lol im joking, good afternoon anyways

[full silo]

has anyone here worked with fine tuning open source text to speech models using custom voice??

[warm forge]

some one have the link to español latino models? pls

[elder willow]

https://discord.com/channels/1159260121998827560/1175430844685484042 and select the tag "spanish"

or you can go to ai hispano, there's more latin spanish models there

[warm forge]

i need a link? or how i can go there

[elder willow]

i don't have the ai hispano server link, you can ask vidal

he maybe has the link

[warm forge]

thanks, but who is Vidal

haha

[elder willow]

engineer of ai hub and member of ai hispano

[warm forge]

how can i talk with him

[elder willow]

leave it for later, i just send him a dm to see if he gives me the link or not

[warm forge]

oh nice, thank you so much

[elder willow]

you're welcome

[elder willow]

i was about to ask ideas for a model

but i will do a model of mendora from the madness disks lol

[dusty atlas]

Generate Image keeps on Waiting to start

[lavish kraken]

hi

[elder willow]

hello

[elder willow]

anyways y'all, don't expect me to stay here for too long xd

[hot verge]

hi people, can someone help me to enhace a video please? :c

[solar torrent]

Imagine talking for a moment and then leave thereafter.

[kindred kelp]

Did you know with gpt search you can find contact information for people?

[fresh shale]

Is there any free AI Lyric Swap?

[valid parrot]

hi

[solar torrent]

Huh. Better not finding the information of someone they don't want you to know.

[light girder]

oooo refinegan seems to sound really cool!

i wonder if we will be able to use it realtime one day

[sacred glade]

Is there like any dogwater mic voice changers?

[covert lake]

What?

[sacred glade]

Any like Voice models that convert your audio to dogwater

A webcam mic

In a nutshell

[night lake]

Grab a bit crusher and crank it

Maybe also increase gain

[iron elk]

hey what's the step to epoch conversion?

[light girder]

btw razer your voices are wesome!!

[night lake]

Thanks

[tranquil lantern]

check ur weights folder

[light girder]

just wondered, do you use your voices on realtime ? If yes how can you make the MOST out of it

[night lake]

Also act it, it’s more convincing if you act like a girl too

[jolly thorn]

привет

https://tenor.com/view/tales-from-the-darkside-hello-chat-funny-discord-1980s-gif-10756244405761972309

[solar torrent]

On W-Okada, you can just lower the extra number down to the lowest number it can reach so the audio can sound in very low quality.

[solar torrent]

Imagine using a voice model that trained from your voice over your actual vocal.

[light girder]

tht'd be CRAZY

[ionic pumice]

mf might be playing gta 7

[bright arch]

Dang, i am stuck in 4

[light girder]

im stuck at 3

[covert lake]

Emulate the 4 on your phone

[fast hollow]

do u guys have ai server for spanish people?

[ionic pumice]

ai hispano ?

[light girder]

true!

[wind mountain]

can someone help me with invoke ai?

[heavy fiber]

free resume builder

ai tool?

[wind mountain]

i want to download stable diffusion model v3.5-large, ive accepted tos or something on huggingface and i got this error: We recommend visiting the repo. The owner may require acceptance of terms in order to download. stabilityai/stable-diffusion-3.5-large

[wind mountain]

figured it out

[north bronze]

What’s the best girl ai voice?

[night lake]

this one: https://discord.com/channels/1159260121998827560/1341216399372062823

[glad nebula]

https://tenor.com/view/jeff-land-shark-marvel-rivals-video-game-gif-10022472975975249

[north marsh]

hi

[mortal bough]

Curious what people are using for ai voiceovers or lipsyncs

[night lake]

https://tenor.com/view/dancing-cat-dancing-kitten-60fps-boogie-cat-gif-24303276

[thin root]

whel

whelp

[paper viper]

does anyone know any repos for adversarial attacks on ai detectors to bypass them

[west burrow]

ok who the fuck is dis

[west burrow]

So, if there’s a classifier slapped on top of the model (like llama guard or anthropic’s constitutional classifier), your best bet is to hit the edges of the datasets that these were trained on

But I always find that the dumbest prompts ever, work

[paper viper]

I found one
https://github.com/zhouying20/HMGC?utm_source=chatgpt.com

I dont know how to use it yet

Ill try and figure it out

i have a shit ton of work due on friday

so

hopefully all goes well

[west burrow]

Ohhh you’re on about those classifiers

[paper viper]

i dont know what a classifier is

[west burrow]

Easiest thing to do is just reword it yourself

[west burrow]

Detector, turns a series of words into an overall class

‘This movie is bad’ -> negative

[paper viper]

doesnt look to hard to set up

it has provided data

with pretty good accuracy

read the paper js now

[rich path]

hello guyz! wassup?

I am currently working on project of making Highly Advance AI Assistant so i need voice model for my Assistant i can't train one myself because i don't have good samples so anyone with Girl voice model please help if uh have anime girl model that would be awesome.\

[covert lake]

Voice overs can be done via voice acting and then use RVC to change voice, or via TTS but won't have emotions

Lip sync can be done by anitalker or face fusion

[covert lake]

You can search rvc ai voice models at:

• #1175430844685484042
• In #🔍│find-models , Do /find with @hidden grotto
• https://weights.gg/ (login required)
• https://huggingface.co/models (but watch out cus in hugging face there arent only rvc ai voice models)
• https://voice-models.com/
• https://thevoicemodels.com/ (for Turkish Models, login required with discord and level 2 on their server)

if there isnt one, you can:

• #1159289738314919936
• #1191429836321849435
• make it yourself with our docs guides https://docs.ai-hub.wtf/essentials/how-to-make-voice-models/

[hidden grotto]

:wave: @covert lake, How can I help?

Available Commands:
• @weights find <query> or /find <query> - Search for RVC Voice Models
• /create - Create an AI Cover
• /image - Generate an Image

[covert lake]

RVC models are STS, not TTS natively unless you use another TTS to make an audio then use it as an input

[rich path]

yes emotions can be done if uh highly train so-vits-fork with more then 100 - 150 good sample with diffrent emotion uh can achive good emotions

[covert lake]

So vits SVC?

That's old asf

It has been replaced by RVC V2 since 2 years

Unless you're talking about another program