🧬│ai-chat

[queen kernel]

Hii

[solar torrent]

You just left the server, so don't expect to see me again.

The last update of this readme metadata for "original" W-Okada was 7 months ago. v.2.0.69 is the latest for Windows.

[polar flax]

meanwhile the deiteris fork was last updated on Dec 2024

and both said versions don't work for RTX 50-series, so you'd have to consider some newer one like vonovox or another tweaked okada fork in our guides

[harsh junco]

oops sorry, i dont really use github that much lol, i think you can tell too

[sharp flare]

can we actually make the AI on weights.gg scream

guess it depends on the model

[polar flax]

use some metal singer model

[sharp flare]

then we can pass it over to my desired model?

[polar flax]

many ordinary models struggle on screaming since they dont have enough of that part in the dataset

[solar torrent]

Yes, it depends on how the voice model can make screaming noises over an audio track. Some can accurately scream, while others simply just make static noises due to its dataset.

[sharp flare]

ahhh

i guess im better off getting its files and training it to scream

thanks for the help

[median cairn]

should I go back making ai covers?

I'm that person who made dio ai cover and reached millions of views by showcasing RVC back 2022-2023

idk if it's worth it uploading on youtube idk doing my hobbies like I used to

with effort

with good mastering and mixing

I wonder how RVC is doing

it's been years

[solar torrent]

Well, it's up to you. Although doing AI covers on YouTbe today can get your channel terminated.

That's too much messages you sent.

[median cairn]

just expressing my feelings about it yk

I just missed this era

this server

nostalgia

[median cairn]

what about japanese songs

should I reupload that kickback dio ai cover?

idk if those record label of that kickback cares enough to strike once I uploaded it?

cuz i deleted it on 2023

[solar torrent]

[median cairn]

cmon I put effort to it as if it's an official audio and it's just a purpose educational only ya know i hope they understand that they pmo

[solar torrent]

Which one of these should I respond to?

[median cairn]

and I wonder where to inference

to an app standalone?

because I used google collab back then

now that i'm back with my life and got gaming pc

i think i can run rvc now

should be faster

I remember only using harvest cuz it's bassy and sounds good

and make sure my input audio isolation vocals only are perfect with no such harmonics

by using that mdx vip model

first of all I research about that particular song if it has an official instrumental

if it does then I'll use it to remove the instrument by overlapping the instrumental only by the original audio

then we got the clean acapella but not perfect and used ultimate vocal remover to clean that off

man

I missed doing all of these

not one of those people who just uneffort without mix and mastering

just straight up using vocal ultimate remover to remove instrumentals

and slapped it into rvc

like they sum tuff

[elder willow]

yee me too @median cairn

[median cairn]

hell yeah

2022-2023 is ai cover era

downside, I hate who made rules about ai generative works but fair enough thats why i stopped uploading

[elder willow]

But look at the bright side

[solar torrent]

Well, if I had to explain, I would have type in lengthy but compact in a message at once and not many different messages explaining about the same "thing" over. While I understand about your excitement over RVC, AI cover and such, I'm just having issues trying to catch the main point from your messages.

[polar flax]

||Sony and UMG have killed it since last year||

[median cairn]

i'm gonna murder them 😄

[solar torrent]

While I haven't uploaded any AI cover music ob any social media as of 2025, I use Weights. When I have no idea which song AI cover I should make, I usually make various voice models to sing BBL Drizzy over and over.

[median cairn]

I use fl studio, audacity, and manually make them sing with final post production quality.

but wow

now there a site that it does it for u

I remember weight site is just used to upload model and download cuz of issues before?

[still sigil]

@oak flower no advertising

[solar torrent]

I use Reaper to mix tracks.

As of 2025, Weights no longer provided a feature where you submit your voice model there to get accepted. But then there's this feature.

[proper finch]

👀

[solar torrent]

Tracks mixing on Weights aren't really that good. It simply applies a raw converted vocal on an extracted instrumental track and that's all. Some of which would sound quiet, and even delayed audio due to RVC issue.

Sending a link without an explanation, or promoting your things are not allowed in this server.

[proper finch]

No worries

[solar torrent]

Thank you.

[minor blade]

ghostwriter????

[solar torrent]

Seeing your message in my direct message box. No, I didn't form Weights.com website, nor I am the CEO of it. What you see in my server nickname is a running gag about this server name, "AI Hub by Weights".

[proper finch]

Thank you!

[half junco]

Does anyone know any free AI courses that has a free certificate. i have been wanting to get into AI development

[echo inlet]

not really familiar with any, but you can opt to go on youtube. there are a bunch there

[proper finch]

AI music kinda dying chat. You think we should bring it back this summer or…?

[stark escarp]

naah let it burn now

soo later itll be back stronger and better than ever

[abstract summit]

why did they do that to weights 🙏

[marble cloak]

Guys, what voice changer app do you use?

[minor blade]

We mostly recommend Deiteris Fork or Vonovox.

For further help go to the #✨│ai-help channel or #1192011222023950368

[marble cloak]

its for free ?

[minor blade]

Yup.

-realtime

[pine acorn]

💻 Local Realtime RVC

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

• Wokada Deiteris Fork

Most suggested. GUIDE

• Original Wokada

ONLY the latest alpha comes close to the Deiteris Fork performance, older versions in youtube tuts are way worse. GUIDE

• **RVC GUI Mainline Realtime**

Unavailable, the guide is outdated and the program is worse compared to the ones above, and much less updated

[green yarrow]

this might be the wrong place to ask but should i learn So-Vits-SVC or RVC if i want to get into ai voice for music

[queen kernel]

RVC is good

This server mainly focused on RVC

[minor blade]

I strongly suggest RVC.

[green yarrow]

I just set up RVC. It's surprisingly easy to use.

[cobalt ivy]

i need a voice model for ab soul

[elder willow]

how do i use gabox's v5 vox it always gives me an error

[daring gust]

هلا

[elder willow]

i hate ai now..

[covert lake]

Was, it won't be anymore

[covert lake]

For help use #1192011222023950368

[covert lake]

You can search rvc ai voice models at:

• https://discord.com/channels/1159260121998827560/1175430844685484042
• In https://discord.com/channels/1159260121998827560/1163592055830880266 , Do /find with @hidden grotto
• https://weights.com/ (login required)
• https://huggingface.co/models (but watch out cus in hugging face there arent only rvc ai voice models)
• Suggested Models for Realtime Voice Changing (Wokada)
• 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:

• make it yourself with our docs guides
• Ask a free request in https://discord.com/channels/1159260121998827560/1159290139609137264
• Be aware that we don't allow any paid comms, so don't fall for any "pay me 20 dollars and i will make the model for you" dm

[hidden grotto]

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

Available Commands:
• @weights find <query> or /find <query> - Search for RVC Voice Models

[covert lake]

#1192011222023950368

[covert lake]

This server is English only

[daring gust]

K

[covert lake]

Huh?

[daring gust]

Yes, is there a bot that understands Arabic that I can add to my server?

@covert lake

[covert lake]

You mean like a Chatbot? We don't even have a Chatbot here

[daring gust]

Ok

[slow palm]

<@&1159293140440723499>

It's in other chats too

[molten acorn]

Hello what local LLM would you recommend for C++ and Python programming and it has to be max 6 months out of date

[chilly lake]

depends on your hardware

Qwen-Coder 2.5

[molten acorn]

Intel

[chilly lake]

or if you got 5090 to run in it then llama3.3_nemotron_super_49b with thinking on

[molten acorn]

What is the knowledge cut off date?

[molten acorn]

iGPU

I dont care about speed that much 2 token/s is enough 32GB ram intel core ultra 7 155H

[chilly lake]

nemotron is like 4-5 months old, qwen coder ~9

[molten acorn]

But the nemotron you specified would be too slow

9 months is too much

[chilly lake]

nemotron :Data Freshness: The pretraining data has a cutoff of 2023 per Meta Llama 3.3 70B:

you wont find anything newer

[molten acorn]

You said its cut off 4-5 months old

[chilly lake]

I did not say that, I said the model was trained 4-5 months ago

[molten acorn]

In how much time will be a model with info max 6 months old available?

[chilly lake]

well, if it takes 6 months to train a model and as much to prepare data.. make a guess

[molten acorn]

in a few years

Well nevermind then which one do you think would be the best no matter the learning cut off date?

[chilly lake]

start with Qwen-Coder 2.5

[molten acorn]

Thanks

[strange summit]

hello is there anyone can help me with RVC

[meager nexus]

Hey, just how significant is 1000 more context window really? By negligibly slowing it down using the original setup (34 -> 32) I was able to get

I was able to go from 16k to 30k stable
honestly, kinda removes the need for deep seek at least for real world use side of things
this thing outperforms deepseek version I can run on mine with only 2k less context

Qwen3 is more accurate then the Deepseek R1 8B

[minor blade]

Elaborate about your issue in the #✨│ai-help channel or #1192011222023950368 one

[craggy heart]

where do i install this?

for voice changer

i cant find how to install it

i need help

[strange salmon]

hii im new here, i started using hailuoa ai to make good ai videos. im actually trying now to learn how to do better prompts

[covert lake]

elaborate in #1192011222023950368

[exotic arrow]

Hello im cris nice to meet everyone

[plain lynx]

Guys I need someone to user test something

[dark jolt]

what voice chnager should i use

[exotic arrow]

Hello i'm have a quewtions if someane like the show américa next top model by tyra banks talking with me a pm

[heavy spindle]

Does anyone have a model called "LyraAI"

?

[exotic arrow]

Let me explain i'm planning make a ai modeing show like américa next top model by tyra banks which one help me

[still sigil]

ai modelling show ☠️

[exotic arrow]

Yes

[still sigil]

https://tenor.com/view/jerry-meme-tom-and-jerry-meme-jerry-points-get-a-load-of-this-guy-this-guy-gif-11384860625931393733

[exotic arrow]

I'm have the Models in sims 4 version

[still sigil]

[polar flax]

https://tenor.com/view/404-not-found-error-20th-century-fox-gif-24907780

[heavy spindle]

oh alr

[exotic arrow]

So if someane like help with My ai modelling show send me a pm

[supple frigate]

This is me

[exotic arrow]

Is someane here like modelling shows like américa next top model by tyra banks talk with me

[elder willow]

some one help

i need some one to give me the file from the video https://www.youtube.com/watch?v=SxdnGxicJOg

pls help

[still sigil]

stop sending the same thing over and over again

[elder willow]

pls help me i need the file from this vid pls https://www.youtube.com/watch?v=SxdnGxicJOg i cant use github

[still sigil]

why cant you use github

[elder willow]

my acc is suspended they havnt gotten back to me for 2 months and alt accs for some reason dont work cus the verification

pls help

[exotic arrow]

im only need help make the ai models & ai photoshoot for my ai modelling show

[still sigil]

so basically you need help with everything

[exotic arrow]

yes

im have sims 4 for turn into ai models

[still sigil]

and thats what #✨│ai-help / #1192011222023950368 is for, not here.

[elder willow]

do you have the file?

[still sigil]

send a message there ONCE and youll get a reply if someone wants to help

[still sigil]

no cuz its all on github

you dont need an account to download tho?

[elder willow]

you cant download it and send it to me

[elder willow]

ye

[still sigil]

yeah

you DONT need an account to download

[elder willow]

no you do

[still sigil]

no you dont.

im not signed in and i can download

[elder willow]

i just showed yu i sent vid

[elder willow]

well i cant so pls do it for me

[still sigil]

i just showd you that you can

[elder willow]

it was cus my acc was still connected to the brave browserr so i had to do it from a dif browser

thanks for no help!

[polar flax]

you dont need an account to download from public github repos

[elder willow]

well i have a suspended acc thats why

and its still connected to my browser i had to use a dif one

[polar flax]

I repeat,

you dont need an account

and it's not like if ur ip is also banned there

[still sigil]

if its suspended just clear cache

[polar flax]

and clear the github cookies

[dark jolt]

anyone got a model called latina egirl

[elder willow]

no bro stop asking

[dark jolt]

why

[dark jolt]

why

i need one

[elder willow]

to bad so sad

[dark jolt]

wihat voice chnager u use?

[elder willow]

gate keeping srry

[dark jolt]

why

is vc client good

[dark jolt]

is rvc the best

[queen kernel]

Maybe

[mint granite]

I need a I need a Latina voice for the Fourth of July

[mint granite]

Why not

[mint granite]

Y

[mint granite]

Bro

[mint granite]

Cuh

[mint granite]

Buddy

[polar flax]

you two..

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

[mint granite]

Fr

[dark jolt]

u use rvc or sum?

bro

no fckng tutorials for rcv

[grave abyss]

@topaz iris

[topaz iris]

I DIDNT EVEN KNOW I HAD THUS INE

ONE

HOW EVEN

[grave abyss]

HAHAHAHA

ITS SO EASY

OKG

OMG

[wise olive]

man wtf happened to weights? now you can’t use any audio over 3 minutes anymore for covers?

[wise olive]

man. guess i’m going back to running rvc locally

i can understand doing this for video or even image generation, but for rvc? i can still run that shit locally on my old laptop with a gtx 1660 in 2 minutes

[chilly lake]

yes you can

even on cpu

[wise olive]

i said that

[bitter hemlock]

I wanna get started with using a voice changer

I used to do it before but long ago now am lost

sm1 pls help

[molten acorn]

Why is it tweaking
https://imgur.com/a/bJoPI5S

its broken

[covert lake]

elaborate in #1192011222023950368

[covert lake]

Weights.com has recently done some changes to the limits due to the extremely long queue crisis there has been recently, along with the expensive they have to run 24/7 with powerful gpus, AI is expensive to run

[covert lake]

you can try Applio or Replay, checking https://docs.aihub.gg or https://tryreplay.io, and ask for help in #1192011222023950368 , ofc that gpu isn't the best so it wont be as fast

[covert lake]

check https://docs.aihub.gg and use #1192011222023950368

[fallen ruin]

https://tenor.com/view/i’am-back-chat-im-back-chat-im-back-chat-gif-1012735009636631433

Chat help

🙏

Where can I read the rules again

❔

[rotund falcon]

guys.. i forgot the voice changer app, do you still remember?

[ashen yacht]

yo how can i use the voichanger in valorant

im trynna not get harrassed

[covert lake]

Hello

[covert lake]

Server guide

[covert lake]

elaborate in #1192011222023950368

[covert lake]

pls elaborate in #1192011222023950368

[covert lake]

Do people really harass girls for playing games?

[ashen yacht]

yeah 90% of my games

or being creepy and sexual comments

🥀

[night lake]

actually yea

so much

[covert lake]

Oh wtf, I'm sorry for you, if you make a post in #1192011222023950368 , and tell your PC GPU and operative system, helpers can help you up

[covert lake]

That's sad

[covert lake]

It's just a game not tinder 😭

[night lake]

guys being weird online is like the big reason e boy voice trolling doesnt work, girls will just ignore/not even be in voice/mute them

[molten acorn]

Hello why is it tweaking
https://imgur.com/a/bJoPI5S

Qwen 2.5 coder btw

32B

Q6_K

[slow mural]

Hi, I'm looking for someone who can professionally clone a voice using RVC. I’ll pay $25 via PayPal.

[fallen ruin]

Thanks 👍

[tender ibex]

who best ai voice for voice changer

[digital coyote]

yo can i use this as a main mic?

if i change

[covert lake]

This is a General AI Server, we won't be focused on voices anymore

Elaborate **in #1192011222023950368 **:

• your PC GPU
• your operative system
• what you want to do
• what tutorial link are you using
• a screenshot of the program

[covert lake]

You can search rvc ai voice models at:

• https://discord.com/channels/1159260121998827560/1175430844685484042
• In https://discord.com/channels/1159260121998827560/1163592055830880266 , Do /find with @hidden grotto
• https://weights.com/ (login required)
• https://huggingface.co/models (but watch out cus in hugging face there arent only rvc ai voice models)
• Suggested Models for Realtime Voice Changing (Wokada)
• 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:

• make it yourself with our docs guides
• Ask a free request in https://discord.com/channels/1159260121998827560/1159290139609137264
• Be aware that we don't allow any paid comms, so don't fall for any "pay me 20 dollars and i will make the model for you" dm

[hidden grotto]

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

Available Commands:
• @weights find <query> or /find <query> - Search for RVC Voice Models

[river verge]

DIlly ding, dilly dong! Two new RegalHyperus drum models just released!
Smash and Grab & Musician (Drum models no. 611 & 612)

[polar flax]

most public chats are toxic, including moba games

usually a safer way is to use discord vc with your trusted friends while playing

[ripe adder]

Hey

[tight light]

Lowkey somebody should make an AI voice of the main manager in squid games. yk the guy who announces every event and stuff. hes badass af.

[covert lake]

You can search rvc ai voice models at:

• https://discord.com/channels/1159260121998827560/1175430844685484042
• In https://discord.com/channels/1159260121998827560/1163592055830880266 , Do /find with @hidden grotto
• https://weights.com/ (login required)
• https://huggingface.co/models (but watch out cus in hugging face there arent only rvc ai voice models)
• Suggested Models for Realtime Voice Changing (Wokada)
• 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:

• make it yourself with our docs guides
• Ask a free request in https://discord.com/channels/1159260121998827560/1159290139609137264
• Be aware that we don't allow any paid comms, so don't fall for any "pay me 20 dollars and i will make the model for you" dm

[hidden grotto]

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

Available Commands:
• @weights find <query> or /find <query> - Search for RVC Voice Models

[meager nexus]

is there any benifit for being able to load 2 llm models at the same time? I have managed to tweak it so now instead of 27 t/s at 15gb vram, I can get 23 t/s at 8gb vram.
I'm using lmstudio btw

I have roughly 8gb of vram to run a small model on of some sort.

[chilly lake]

1 model + big context is better

[meager nexus]

I'm already pretty much maxed the context of this model. that's why I'm trying to optimize other aspects now

[chilly lake]

use better quantization then

[meager nexus]

fair point. currently running it at quant q4_k_m. I got 16gb vram. what you recommend?

I would rather not use KV cache If I can help it since I want it to be as accurate as possible

[chilly lake]

q6 or q8

or a model with more paramters

b12 instead of b8

[meager nexus]

is there one better than qwen3 30b MoE?

[chilly lake]

depends on the purpose

[meager nexus]

academic use reasoning model

I want "free chatgpt pro or gemini pro" that can run on 16 consumer card essentially

the deepseek destilled sucked hard so not gonna try that again

[chilly lake]

you can ask in LM studio discord, there are plenty of people

something good with reasoning may require more than 16GB

[meager nexus]

there's a lmstudio server?

[meager nexus]

I'm currently offloading atm

seems to work fine since I can get 27 t/s at 15 gb usage with 30k context

[meager nexus]

oh, never paid too much attention lol

[meager nexus]

I asked gemini pro to explain the difference with quant for me using audiophile term and it helps a ton to figure out how they work

Prompt: Im an audiophile, compare these file as if there were compression type such as flac vs mp3 for example regarding llm quant

Q8_0: The FLAC of LLMs

• Audio Analogy: FLAC (Free Lossless Audio Codec)

Q6_K: The High-Bitrate AAC/MP3

• Audio Analogy: 320 kbps MP3 or Apple's 256 kbps AAC

Q4_K_M: The Standard MP3

• Audio Analogy: 128-192 kbps MP3

this makes, a ton of sense to me now

[chilly lake]

i would not go below q4

[covert lake]

Q0.5

[meager nexus]

If my goal is accuracy of the model, I should choose q6, correct? I can run q4_k_m at 8.5 gb vram on my 16gb vram card.

[meager nexus]

that's the speed at 23 t/s

[chilly lake]

then try q6

[meager nexus]

oh btw, how is Phi 4 reasoning plus? My pc can seem to easily run the Q8 version

the model ranked higher in gpqa score at 68.9% vs the 65.8% of qwen3 30b a3b

I was looking for Magistral Medium but can't seem to find the model

that one has 70.8% in gpqa

all the other "better" locally runable model was stuff that you need a lot more vram for

[supple frigate]

Is anyone good with editing discord messages with nitro emojis for announcements

[dawn temple]

I can help :)

But I keep my announcements minimal

[supple frigate]

Yes

Can u check dm

[dawn blade]

help

RuntimeError: CUDA error: no kernel image is available for execution on the device
CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
For debugging consider passing CUDA_LAUNCH_BLOCKING=1.
Compile with TORCH_USE_CUDA_DSA to enable device-side assertions.

[chilly lake]

the answer depends on what you're trying to do.. 1) out of VRAM 2) 5000 series gpu?

[polar flax]

it looks like an outdated program or unsupported GPU

(minimum GTX 10-series requirement or RTX cards recommended)

RTX 50-series would need installing latest pytorch and cuda 12.8

[quiet shale]

can someone help me my rvc won't connect to discord

i already download vb audio driver

[polar flax]

vb audio is crappy

[quiet shale]

so what do i use?

so i can use my rvc to discord and others?

[quiet shale]

What should i use?

[high ermine]

Anyone wanna have a conversation? Im pretty intelligent and ai facinates me in the most brilliant ways. However nobody I know wants to talk about its true potential that I visualize. I even built my own little engine that can click around scroll etc. it could also open apps, open google and youtube and search something. And more but yea Idk anyone to talk to about this type of stuff!

[atomic marsh]

@toxic lance

(:

[toxic lance]

: )

[opal sun]

.

[atomic marsh]

(:

[vagrant void]

can someone give me a link to a colab where you can train models (that isn't applio)?

[ripe adder]

try tensorflow model training walktrough

[covert lake]

use https://discord.com/channels/1159260121998827560/1384511238977097858

[covert lake]

elaborate your pc gpu and what you want to do in #1192011222023950368

[high ermine]

Yo anyone on rn?

I really need someone to talk to about ai.

[tired flume]

ai

[deep basin]

I was in this server a couple years ago and I'm rejoining cuz I got back into messing with RVC v2

what does the "by Weights" appended to the name mean? When I was here it was just AI Hub

I saw someone else had sent an invite to the Weights server but I still don't know what it/they is/are

[covert lake]

https://Weights.com is a site we partnered with, which is a cloud service that lets people use AI, such as covers, making models, make images, videos, and much more as much easy as possible

[covert lake]

also, there has been pretty few improvements, the original devs left RVC to rot, RVC is just being experimented around by some of our engineers, but rvc v3 will NEVER come out

[deep basin]

damn

that explains why it hasn't been updated in 2 years

[covert lake]

there are more updated forks, like Applio, but it's mostly just some easier UI and performance improvements

also, we will be a GENERAL AI Server, and will stop focusing on RVC

TL;DR: RVC is dead till another AI replaces it in maybe some years

[deep basin]

a shame

[covert lake]

we will be actually ai hub, and not RVC hub

[deep basin]

though even in its current state it's powerful enough for whatever shit I make messing around with it so I guess it's ok

btw I wanted to ask if there's an updated tutorial for installing tensorboard into RVC

I had one bookmarked from a long time ago but the files are dead links now

[covert lake]

elaborate your pc gpu, operative system, and what you want to do in #1192011222023950368

[covert lake]

yup, dont trust any rvc tutorial outside this server, they are 99.9% outdated

[deep basin]

nah it was the one made by kalomaze

[covert lake]

that one is so outdated that he isn't even apart of the server anymore

he's there but not active since 2 years

[deep basin]

💔

[covert lake]

many staffers changed

many people changed

nothing is the same basically, except RVC quality lol

[covert lake]

last message was over a year ago

[tired flume]

haven't seen kalo in a long time

[covert lake]

and you prob wont again

[tired flume]

why

[covert lake]

people got tired of RVC

[deep basin]

I was never very active here so I don't think I talked to him much but still a shame, his tutorials were the only way I figured anything out in the beginning

[covert lake]

he does LLM stuff now

[tired flume]

good for him

[covert lake]

he doesn't even use RVC anymore

i mean rvc doesn't even have that many people using it daily than like 20 people

@deep basin btw if you're going to make an ai cover, be aware that you can risk getting your youtube account down now

i deleted all my ai covers after getting 2 strikes in less than a day

[deep basin]

I don't upload anything I make thankfully

[covert lake]

alr make an help post so i can help you up then

[deep basin]

Nah it's fine, I found the Tensorboard install guide on the AI Hubs docs website

So I'll just follow that

I'll also look into installing Applio

[tired flume]

i know why

RVC v3

wasn't

released

[lavish fractal]

hey guys i was wondering if anyone knew where exactly all of these ai influencers go to get such high quality images?

[grand python]

^ Honestly, just be like me and either disguise it as a meme on Reddit or just share it with a few friends.

The subreddit I'm in is about to ban AI-art anyway, but there might be an exception for AI-assisted posts.

[deep basin]

I have always only shared with friends

I don't use reddit and the only subreddit I look at I'm shadowbanned from for low karma (because my account is entirely unused)

I posted a shitty cover meme I made in a server though and it got 3 fire reacts and that's all I need

[grand python]

Yeah, the best use for AI covers is personalization for playlists.

I have over 1000 covers in one folder and I just put them on mix when I'm doing work or chilling.

[thick obsidian]

forgot where the download is

where is it

thank you

[covert lake]

This is a General AI Server, we won't be focused on voices anymore

Elaborate **in #1192011222023950368 **:

• your PC GPU
• your operative system
• what you want to do
• what tutorial link are you using
• a screenshot of the program

[covert lake]

why even ban ai art

that's just making reddit's job easier

[grand python]

¯_(ツ)_/¯

[covert lake]

if only people could fucking read and know that reddit sells data to ai lol

but people blindy agree to whatever contract they see nowdays

[thick obsidian]

sure

[thick obsidian]

yes, i apologize simply confused

[digital elk]

have france voice?

[solar torrent]

Try find one in #1175430844685484042.

[night lake]

ah yes the country

as a voice model

[digital elk]

i dont finds french voice

[solar torrent]

"France" refers to a country in Europe, so then there is the word "French" which refers to a person who is French.

[solar torrent]

Do you mean this?

[digital elk]

oooooh

you are so intelligent

thank you miss

i love you

love me too

please

kill

kiss

[limber thorn]

hey guys

what’s up

[supple warren]

Hey guys question

What websites have some decent voice model library?

[mellow crag]

can some tell me how to fix this

@paper blaze @hidden pasture

[chilly lake]

fairseq does not work with python 3.11 that is colab

[mellow crag]

so how i can fix that

[chilly lake]

you can't. In applio we have moved to a different embedder

[supple warren]

What';s the best program to use to add reverb and such to RVC?

Voicemeter banana?

[hasty mist]

why does the bat close when i open

[chilly lake]

what .bat?

[hasty mist]

start_http.bat

[chilly lake]

that's outdated software

-rt

[pine acorn]

💻 Local Realtime RVC

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

• Wokada Deiteris Fork

Most suggested. GUIDE

• Original Wokada

ONLY the latest alpha comes close to the Deiteris Fork performance, older versions in youtube tuts are way worse. GUIDE

• **RVC GUI Mainline Realtime**

Unavailable, the guide is outdated and the program is worse compared to the ones above, and much less updated

[coral marsh]

is there a normal sounding girl voice?

[minor blade]

XD

[main tangle]

gayz plz any help me i am new which is best tool convert voice

[elder willow]

So is it better to make YouTube videos about ChatGPT?

[main tangle]

I AM CREATING SONG VOCAL TRAIN PLZ GUIDE ME

FREE OF COST TELL ME

[glossy atlas]

Has anyone ever used weights n biases / wandb?

[covert lake]

You can search rvc ai voice models at:

• https://discord.com/channels/1159260121998827560/1175430844685484042
• In https://discord.com/channels/1159260121998827560/1163592055830880266 , Do /find with @hidden grotto
• https://weights.com/ (login required)
• https://huggingface.co/models (but watch out cus in hugging face there arent only rvc ai voice models)
• Suggested Models for Realtime Voice Changing (Wokada)
• 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:

• make it yourself with our docs guides
• Ask a free request in https://discord.com/channels/1159260121998827560/1159290139609137264
• Be aware that we don't allow any paid comms, so don't fall for any "pay me 20 dollars and i will make the model for you" dm

[hidden grotto]

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

Available Commands:
• @weights find <query> or /find <query> - Search for RVC Voice Models

[covert lake]

use #1192011222023950368 and elaborate pls

[covert lake]

no need to ping helpers, also this is the wrong channel

[covert lake]

that's from original wokada, DONT TRUST VIDEO TUTORIALS THEY ARE ALL OLD FOR RVC AND WOKADA, the other helper already gave you wokada deiteris fork which is better, next time elaborate in #1192011222023950368

[covert lake]

This is a General AI Server, we won't be focused on voices anymore

Elaborate in #1192011222023950368 :

• your PC GPU
• your operative system
• what you want to do
• what tutorial link are you using
• a screenshot of the program

[covert lake]

what?

you can make any yt video u want

if u wanna talk about chagpt, sure

[elder willow]

I mean, is it better to make YouTube videos about ChatGPT because the AI covers get taken down

[covert lake]

biases?

[covert lake]

if you want to, sure you can do that, it wont get taken down

[elder willow]

Like make a video on how to translate with ChatGPT?

And how to use it to help learn languages?

[covert lake]

maybe, its ur choice

[elder willow]

Should I feel bad that I use AI, I like using AI and chatting with ChatGPT, but some people don't like it.

I like asking it to translate in certain languages.

[slim lintel]

hi can someone help me with a voicechanger

[covert lake]

elaborate in #1192011222023950368

[simple viper]

how do i set up the voicechanger in roblox?

[warm ivy]

Help

[silver sentinel]

u can use third-party apps like voicemod

[solar torrent]

Don't. There's a better realtime voice changer, which is W-Okada.

[finite grove]

ffd

[polar flax]

anyone pls don't fall for sponsored garbage like that and voice.ai

[surreal trail]

Can anyone help when I open the start_http the native voice client dosent show

am i doing something wrong?

<@&1159293204038955078>

[young yacht]

Is there a Google collab link for Training models?

[gray ridge]

@celest cosmos Kannst du mir helfen auf korrekt

[copper pollen]

hey I want model RVC model and is artist name is a mc stan is a Indian rapper I want mc stan Indian rapper mc stan voice model

[copper pollen]

plzzz idk how to make how to get plz give me 💚

[polar flax]

try searching in https://discord.com/channels/1159260121998827560/1175430844685484042 or anywhere over the internet

otherwise make a thread in https://discord.com/channels/1159260121998827560/1159289738314919936

[tired girder]

Is anyone interested in AI-generated game creation?

[gray ridge]

@polar flaxcan u help me?

[summer pecan]

Check Voisi.

[summer pecan]

I think you should look at the value than what others think. There are tools that translate better. ChatGPT may be limited when it comes to quality translation.

[polar flax]

there are machine translators like DeepL

or you can compare with other LLMs

[polar flax]

it's not so easy to do, perhaps with Duolingo or something

[shell valley]

hello yall im new to this server. But i want to start a project for an A.I model that can learn and adapt

and im also combining it with an chat gpt model that they allow people to use in developer thingy

so i can use that as a base and use that as the starting point

and let it slowly learn

@icy kindle You said convernsation or chatbox?

because im working on an chat box as we speak

[icy kindle]

i said voice conversion

[shell valley]

chat gpt

they can speak within that call thing

and its free

but it gets weird at times

[icy kindle]

idk im trying to convert my voice and this is https://huggingface.co/spaces/r3gm/RVC_HFv2 is shwoing error

showing*

it worked yesterday

now it doesnt

[shell valley]

did you try to refresh or close teh tab and open it back up or restart the device?

[icy kindle]

yes lemme shwo the ss of the rror

show*

wait there is no option to share image

[polar flax]

go post a thread in https://discord.com/channels/1159260121998827560/1192011222023950368

[hoary parcel]

hi. is there a way not to show the cmd propt when opening VoiceAI?

[solar torrent]

There are RVC Google Colab notebooks, but this is not where you ask for this. Type "-colab" in #🤖│bots and you'll see links there.

Don't send a link without explanation or promote your link. Delete this.

[solar torrent]

If you're looking for RVC or W-Okada, go to #✨│ai-help or #1192011222023950368 and explain about your issue there.

I see that. Don't spam an image here.

[covert lake]

elaborate in #1192011222023950368

[covert lake]

elaborate in #1192011222023950368

[covert lake]

elaborate in #1192011222023950368

[solar torrent]

Please speak English.

[covert lake]

elaborate in #1192011222023950368

[covert lake]

You can search rvc ai voice models at:

• https://discord.com/channels/1159260121998827560/1175430844685484042
• In https://discord.com/channels/1159260121998827560/1163592055830880266 , Do /find with @hidden grotto
• https://weights.com/ (login required)
• https://huggingface.co/models (but watch out cus in hugging face there arent only rvc ai voice models)
• Suggested Models for Realtime Voice Changing (Wokada)
• 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:

• make it yourself with our docs guides
• Ask a free request in https://discord.com/channels/1159260121998827560/1159290139609137264
• Be aware that we don't allow any paid comms, so don't fall for any "pay me 20 dollars and i will make the model for you" dm

[hidden grotto]

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

Available Commands:
• @weights find <query> or /find <query> - Search for RVC Voice Models

[covert lake]

elaborate in #1192011222023950368

promos aint allowed

[cobalt lynx]

whats the best ai to use

[cloud hornet]

@little abyss 😜

[sage steeple]

i am racism

[solar torrent]

Good for you.

[sage steeple]

i am not racism\\

[little abyss]

😑

[sage steeple]

can someone join me ?

[solar torrent]

No. Promoting your server is not allowed here.

[polar crag]

the voice changer doesn't working on my pc

[solar torrent]

If you're looking for W-Okada, go to #✨│ai-help or #1192011222023950368 and explain about your issue there.

[plush pawn]

starting out my AI research and want to know as much as I can. anyone know a good place to start?

[elder willow]

Google Translate seems better at most things than DeepL does nowadays. Google translate knows gumimaci means gummy bear in Hungarian

[gray ridge]

the owner is german..

[forest coral]

wer deutsch?

[plush pawn]

anyone know?

[rigid bluff]

What are the best setting for the girl ai voice changer

[boreal ice]

Can anyone give me the latest version of wokada?

[flat crown]

hellooo everyone

guys i need a little help

please if you study anything related to AI dm me i would really appreciate it

[simple valve]

I have a question how do I use an existing voice model and put it on a voice over?

like prerecorded voice or song or something

[plush pawn]

does ANYONE know where to start on learning AI

[chilly lake]

you can start on youtube, there are plenty of videos by the friendly indian instructors

[narrow trail]

anyone got a ian model

[worn dagger]

кто русс?

[still sigil]

https://discord.com/channels/1159260121998827560/1159346439424573440

[hushed talon]

hello

[atomic pilot]

hi

[rustic jetty]

Hello all

[atomic pilot]

hru

[rustic jetty]

Not bad bro, yourself?

[heady jewel]

pls help i need, on #✨│ai-help 😭👐

[atomic pilot]

I'm the same too

[brittle sorrel]

how we can train a model and later make a cover? i need help i dont see any colab or something working

[polar flax]

[rustic jetty]

Training a model for cover songs can be hard, especially if the available Colab notebooks aren’t working well or are outdated. It depends on what kind of model you want to train—like for voice cloning, style transfer, or full song generation.

[tranquil hollow]

Does anyone know why you can't even train Lora image models in Weights anymore?

[rustic jetty]

Gradient checkpointing can block LoRA updates—try turning it off.
VRAM limits can cause crashes—use smaller models or batches.
Weights might have platform restrictions on LoRA training.

Check their docs or support, and if it keeps failing, try other platforms or local setups.

[tranquil hollow]

I will try civitai because I have tried 5 times before and it won't let me.

[rustic jetty]

If you’re having trouble training LoRA models on Weights or Civitai, try double-checking that your model and dataset are compatible with the platform. Sometimes clearing your cache or storage helps if uploads keep failing. Make sure all your tools, scripts, and libraries are up to date too. Don’t hesitate to reach out to the community on forums or Discord—they often have quick fixes. And always remember to back up your data and weights regularly.

[brittle sorrel]

I just want to create the voice and then do a cover

[tranquil hollow]

I tried to train with weights, but it says I reached the weekly training limit, and I didn't train until I hadn't done it for 12 days.

[rustic jetty]

To create a custom voice model, you’d need to train a voice cloning or TTS (text-to-speech) model with sample audio of the voice you want. There are some tools like Coqui TTS or VITS that people use for this. After training, you can generate new singing or speaking audio to do your cover.

If you’re new to this, starting with open-source projects on GitHub or some Google Colab notebooks for voice cloning might help you get a feel for the process.

[rustic jetty]

Sometimes these platforms have strict limits on how much you can train within a week, and the counters don’t always reset exactly when you expect. It might be worth reaching out to their support or checking if there’s a way to manually reset your training quota. Also, some people try using different accounts or platforms to get around those limits, but just be careful with the terms of service.

[brittle sorrel]

can u explain better?

[tranquil hollow]

Where can I contact support?

I just tried with another account but it gives me the same thing, and I barely create it

[rustic jetty]

Google their support email

[sonic needle]

Hello everyone

[native pecan]

hmm

how did i get pinged for that last announcement when i dont have the role, discord be trippin'

[vital venture]

istg doing anything AI/RVC is so cancer on AMD

[rustic jetty]

Anyone interested in AGI?

[scarlet drift]

Hewooo~

[polar flax]

hi

[scarlet drift]

A friend i think got banned from here... can someone un ban them somehow...?
If i remember correctly, it was because they got hacked by someone and the same "hacker" SENT NSFW INVITES to all their dms and servers... which got them banned from a few

[polar flax]

send a ticket to @serene sluice

[scarlet drift]

thank you!

[hallow iron]

anyone know where i can find ai converter like kits?

[round dagger]

Still rvc is best ai cover tool?

It’s 2023 tool looks old

[lavish oasis]

Hello

Please help me, I have a HackTool:Win32/AutoKMS!MSR virus

[sick berry]

malware bytes

[still sigil]

burn your pc and buy a new one

[opal badge]

how come the voice changer for mac is so slow

[copper pollen]

anyone make model for me and is artist name is MCStan Is indian best and no.1 rapper

[elder willow]

Hey so I've trained a model. My only question is what is the best fork of RVC? I heard mangio is good. Anything I should know?

[astral nova]

umm in the rvc models are the tags working?

[polar flax]

wdym tags?

the ones in the posts in https://discord.com/channels/1159260121998827560/1175430844685484042 aren't applied to their model itself

[astral nova]

ooh

oh

wait

[polar flax]

Windows just hates it

or move to Linux if you wish

[timber fox]

my app isnt loading

[chilly lake]

seems like you have a windows activator instead 🙂

[worldly bloom]

hello

[gray ridge]

@polar flaxcan u connect me with owner

[polar flax]

everyone is not the owner

so could you stop pinging?

[gray ridge]

are you iq losses?

[gray ridge]

someone did Create this server o-o

[polar flax]

ye lemme play dumb

[gray ridge]

i see u dumb lol

[polar flax]

cuz I don't know whose "owner" that speaks german

[gray ridge]

So how did you became mod

if you dont know who is it

[polar flax]

you could even report to demote my mod role

[gray ridge]

?

[polar flax]

1 + 1 = 3

okay?

[gray ridge]

ay you fucking idiot How the bloody hell can u tell me you dont know the owner if u became mod role

Are u okay?

[polar flax]

no im not okay

[polar flax]

btw if you need to talk to him

just contact

not through anyone else including me

[still sigil]

does she speak german?

[chilly lake]

Looks like someone who grew up without fear of recounting their leftover teeth.

[gray ridge]

jaa

[still sigil]

incorrect cuz the server owner is a he

[gray ridge]

he speaks german to ik him

[still sigil]

who is it then

if you know him then youd be able to ping him

[gray ridge]

i was pinging him

yesterday

[still sigil]

miizu is not the server owner

nothing to say?

typical

[chilly lake]

thot pfp

[gray ridge]

i didnt mean him

[still sigil]

ping him again then

the owner

[weary flare]

Hi guys

Is it better if I download v2 or stay on v1

[chilly lake]

v1 what

[lethal vault]

Guys, I'm trying to train a voice model, but even after two attempts, the output isn’t turning out as expected. The audio itself is spot on, but there's a clicking sound at the end of each sentence when the person finishes speaking. I'm using a clean 2-hour long audio sample, and I’ve confirmed that it has no background noise or artifacts. can someone please help me with this issue

[polar flax]

• try doing inference without enabling "Split Audio" in Applio
• for the model dataset, try using less aggressive noise gate and/or truncate silence

• and make sure there's no much popping in the dataset

[lethal vault]

the data set is clean no popping sound or anything I even tried using a different voice it was also over 2hr but same thing after the person is done speaking there is a click at the end for some reason

All i did with the audio was run it through audacity to convert it into mono should i clean it more using some other softwares

[chilly lake]

show your input audio

[wintry coral]

Is anyone having trouble with rvc as of right now?

My res keeps going up and down, and the audio repeatedly cuts in and out.

I have a RTX 5070 twelve Gb graphics card.

[chilly lake]

realtime voice changer?

[wintry coral]

Yes,

[chilly lake]

https://rentry.co/forkvoicechangerguide download 5000-series version

[wintry coral]

It says index not found on my device.

[wintry coral]

?

[solar torrent]

Fork W-Okada RTX 5000 GPU is the only known W-Okada that can run on RTX 50 series GPU. Also, for issues about W-Okada, go to #✨│ai-help or #1192011222023950368.

[solar torrent]

RVC voice model version (RVC v1 or RVC v2) or the "original" W-Okada versions (MMVCServerSIO_win_onnxgpu-cuda_v.1.5.3.18a) or anything else?

[spiral wasp]

Quick question: What's the best f0 est.? for okada

[solar torrent]

You know, you look stupid if you get angry enough to call a mod an idiot. While neither a moderator is the server owner (I'm not sure which server owner you're looking for), there are admins here. You can at least use the modmail account (AI Hub Modmail) if you have any issue in this server, whether someone is impersonating you or spamming.

[solar torrent]

Rmvpe.

[spiral wasp]

why's that? 🥺

[solar torrent]

Rmvpe for W-Okada with NVIDIA GPU, any other GPU (AMD/Intel) is rmvpe_onnx. Simply.

[fiery birch]

Random can somebody please tell me why there are so many tech layoffs.Yet accounting seems to be more stable right now.I know things in the future will eventually be taken by ai but would like to know why software engineering jobs seems to be more impacted by ai then accounting for example.

[sterile tundra]

Test

[chilly lake]

laying off people from USA, hiring people on H1-B / remotely in Bangalor

[split ingot]

hi guys tell me how to download voice changer

[fringe vessel]

anyone here not using the exitlag in the gift inventory?

[mortal kelp]

who can help , i need cloning voice

[queen kernel]

Ask in #✨│ai-help and ping helpers

[limber coral]

how i make an ai cover?

[dusk urchin]

Whats perplexity for?

Their WhatsApp AI seems to be like chatgpt, but the app itself seems to be for investigation or research

[lyric sierra]

is there a fix to it sounding robotic or is that just the model itselfs issue

[nimble panther]

/generate voice: جورج وسوف audio: (ارفع أغنية "الصبر جميل")

[minor blade]

It's either your settings on W-Okada or a problem on the model itself.

Use the #✨│ai-help or #1192011222023950368 channels for further help

[lyric sierra]

it feels like half of these models suck so it might be my settings or good sounding models are just rare

[night lake]

good sounding models are rare

but can also be your settings

[lyric sierra]

im running 500ms 2 chunk with a 4060 idk if thats good or not

[fleet linden]

Hey can someone help me get the voice changer up and working

[hot tusk]

@jameso.0 yes I can help you with that

[solar gyro]

Is someone online

[weary flare]

Ya that

[radiant hearth]

sorry if this isnt the right place to ask, but ive tried everything it feels like, i got this W-Okada voice changer software that works fine in OBS etc but i cannot get it to work in fivem/gta, if anyone has any ideas id be grateful

[dusty ingot]

BRO

the voices arent working for me

@radiant hearth

@weary flare

@solar gyro

@hot tusk

@lyric sierra

@night lake

[still sigil]

[dusty ingot]

bro

help

me

@still sigil

help me out

help

help

help

help

lpawaw

w

w

w

[alpine ore]

can someone help me with w okawa its really laggy

[weary flare]

Dm me

[weary flare]

Are u run on cpu or gpu, what ur spec

[alpine ore]

im on

gpu

my gpu is

radeon amd

[weary flare]

Which

[alpine ore]

5 5600

[weary flare]

Wuh

Gpu?

[alpine ore]

oh

mb

[weary flare]

U put cpu

[alpine ore]

XFX Speedster SWFT 210 Core Radeon RX 7600 8 GB Video Card
XFX Speedster SWFT 210 Core Radeon RX 7600 8 GB Video Card

[weary flare]

Lemme do sum research rq

Brb

[alpine ore]

oh okay

[weary flare]

Xfx speedster swft 210 core Radeon 7600 8gb

[polar flax]

if you don't see ur gpu detected, make sure you've downloaded the correct AMD version

https://docs.aihub.gg/rvc-voice-changer/local/deiteris-w-okada-fork/#download-amd-intel-and-cpu-on-windows

[alpine ore]

i doo

[polar flax]

just switch to the gpu, not cpu

[alpine ore]

i did

[polar flax]

then?

[alpine ore]

it was always

[weary flare]

How laggy?

[alpine ore]

like

un hearable

[polar flax]

try turning up chunk?

[alpine ore]

it sounds like

[weary flare]

U can't hear?

[alpine ore]

ah

for a

[weary flare]

It's choppy?

[alpine ore]

.0000000000000001 secend

really choppy

[polar flax]

pls make a thread in https://discord.com/channels/1159260121998827560/1192011222023950368 to continue

[alpine ore]

done

@weary flare

are you still there??

[weary flare]

Ya

[digital thicket]

Hello

Anyone here using voice.ai pro version? I wanna ask if it can change my voice in a meeting

If somebody can tell me this I would be really grateful 😀

[polar flax]

still using that garbage? try considering this free alternative

https://docs.aihub.gg/rvc-voice-changer/local/deiteris-w-okada-fork/

[digital thicket]

So my main use case is sounding American

I'm an English teacher and I teach Chinese people English online

Sometimes they argue that my accent is the problem

Will it fix that?

[digital thicket]

Thank you can it work on zoom?

[polar flax]

just find and use some American speaker model

[digital thicket]

Thank you brooo

It's a Live voice changing tool right?

@polar flax Sorry for mentioning again but it needs a gpu right? I'm using aa laptop with 128mb integrated one 💀

[polar flax]

the spec requirements clearly says RTX 2060 as minimum recommended

(or laptop 3050)

[digital thicket]

Is there like a voice changer I can use that has a monthly subcription of like 10-15$

On my laptop smoothly

i7 4th gen 16gb ram

[still sigil]

4th gen

[polar flax]

this is the cloud alternative https://docs.aihub.gg/rvc-voice-changer/cloud/w-okada-kaggle/

[digital thicket]

Thanks

[fresh harness]

chat

dy guys know any good AIs which make sites on prompt?

[fiery zealot]

Hello
Does anyone have experience with the SMC trading robot? If you do, I would really appreciate a bit of help or guidance. Please kindly reach out if you're familiar with it.

Thank you in advance!

[neat tree]

can you check this video and tell me which ai cloning software the person is using to get such accurate results

https://www.instagram.com/reel/DLfgNrOSu5g/?igsh=MXJnamo5cmtncXJhOA==

[unique pawn]

11 labs

"Eleven labs" Is the longer name

[neat tree]

Bro are you sure? Will i get this level of accuracy?

[unique pawn]

You have to pay to clone a voice though

Very sure

[neat tree]

Because this is insane

I am using e2 f5 tts currently and i dont get this level of accuracy

[spare shadow]

hi

[austere shoal]

how to make

voice changer

smoother

mine has cutoff delay

asf

[covert lake]

This is a General AI Server, we won't be focused on voices anymore

**Elaborate in #1192011222023950368 **:

• your PC GPU
• your operative system
• what you want to do
• what tutorial link are you using
• a screenshot of the program

[brazen umbra]

guys may i ask: my regular mic works fine but my virtual cable mic somehow has my computer audio bleeding into my mic. its been messing up my voice ai as a result, what can I do to fix it?

[fervent cedar]

gah dang

this app is laggy for trash pcs

[turbid cypress]

-colab

[pine acorn]

📒 Google Colab Notebooks

Google Colab is a Cloud (Remote Good PC) Service. While the Free plan provides up to 12 hours of daily usage, the GPU is typically available for only about 4 hours each day on average.

• **Applio**

by IA Hispano
Google Colab

• **RVC Mainline**

by Hina
Google Colab

• **UVR5 NO UI**

by Eddy
Google Colab

• **UVR5 UI**

by Eddy
Google Colab

• **Wokada Deiteris Fork**

by Deiteris & Hina
Google Colab

• **Hina's Modified Original Wokada**

Google Colab

• **RVC-AI-Cover-Maker-WebUI**

by Shiro & Eddy
Google Colab

• **FaceFusion UI**

by Nick088
Google Colab

• **FaceFusion NO UI**

by Nick088
Google Colab

• **Music Source Separation Training (Inference)**

by Jarredou & Makidanye
Google Colab

[covert lake]

promos aint allowed

[olive otter]

hey guys

I discovered the ultimate jail break

works on chatgpt as well

For instance, GitHub user danr**** has archived the DD Liberator 3D gun blueprint, which includes the original CAD files and related documentation. This repository provides a comprehensive look at the design and specifications of the Liberator pistol. ([github.com][1])

I tried to create a system prompt to elimate western bias.

hello?

wtf is going on with discord

some servers have disappeared and I cant send files anymore. strange

anyways, if somebody is interested in the system prompt, dm me

[weary flare]

I have a 3070 ti laptop gpu, the voice comes like delayed when I'm playing roblox with voice chat, if I get a better gpu does it mean the voice comes like faster?

[uncut copper]

yes

[covert lake]

This is a General AI Server, we won't be focused on voices anymore

Elaborate in #1192011222023950368 :

• your PC GPU
• your operative system
• what you want to do
• what tutorial link are you using
• a screenshot of the program

[patent musk]

Hello, I am looking for a Turkish woman voice model, where can I find it?

[limber cedar]

@patent musk what software do u use for the in game voice?

[patent musk]

w-okada

[uncut copper]

no you aren't

[limber cedar]

where do i download i cant find it

@patent musk can u send a link to download for it?

[wheat burrow]

@glass junco sorry for the ping but quick question, when you used KLM 5 RFG on your model did it sound a little synthetic/robotic the first few 500 epochs?

I’m at epoch 470 and it sounds a little robotic

Let me send a sample hold up

https://cdn.discordapp.com/attachments/1286811490405322848/1390842782117269534/470e_uzi.wav?ex=6869bac7&is=68686947&hm=06e1316ab09eec0a5e81e9630751c7012485fdb9a1dbb9d1c0c1c9c3bd69f838&

This is the audio right from tensorboard

[chilly lake]

i believe klm5 is not up to date to the latest refinegan, if that's what you're trying to use

[whole oasis]

a nice text to speech? please @ me if any1 responds now

a free one please

[covert lake]

You can search rvc ai voice models at:

• https://discord.com/channels/1159260121998827560/1175430844685484042
• In https://discord.com/channels/1159260121998827560/1163592055830880266 , Do /find with @hidden grotto
• https://weights.com/ (login required)
• https://huggingface.co/models (but watch out cus in hugging face there arent only rvc ai voice models)
• Suggested Models for Realtime Voice Changing (Wokada)
• 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:

• make it yourself with our docs guides
• Ask a free request in https://discord.com/channels/1159260121998827560/1159290139609137264
• Be aware that we don't allow any paid comms, so don't fall for any "pay me 20 dollars and i will make the model for you" dm

[hidden grotto]

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

Available Commands:
• @weights find <query> or /find <query> - Search for RVC Voice Models

[wheat burrow]

im using applio from like

a few months ago

I think February?

[covert lake]

elaborate ur pc gpu and operative system in #1192011222023950368

[wheat burrow]

actually if theres an exact build number i can say it

[covert lake]

tell your pc gpu and operative system in #1192011222023950368

[wheat burrow]

let me find it

[chilly lake]

there were some builds that are compatible, the latest one is not

[whole oasis]

great thanks

[chilly lake]

right now I dont have plans on making a new rfg pretrain

[wheat burrow]

yeah im on one of those compatible builds

i tried codenames fork

it failed to save the files for the epoch

so i just stuck with my applio build that supports rfg

[whole oasis]

i did create the post here, nick088

[jagged dune]

Has anyone come across a more rigourous definition of AGI?

[floral zealot]

Hei guys had some questions, does okada supported for mobile ? I try to run it on google collab but it's keep saying modulenotfound error

[random drift]

Does anyone else feel disguisted when they see ai images? Im not trying to hate on it but whenever I see an ai generated image I feel disguisted like I want to crawl out of my skin

[polar flax]

maybe for earlier SD and other models
but nowdays there are quite many AI images being less distinguishable until you carefully peep the details

[chilly lake]

#🏙│ai-images message

even this?

[still sigil]

@small grotto no advertising

[scenic zodiac]

Guys do anyone knows which AI will be best for free video generation for mobile n for PC?

[clever arrow]

Guess what

… chicken butt

[covert lake]

No, wokada will never work for mobile, even if you use cloud (remote good PC), it won't work in other programs because phones lack of a VAC

A VAC (Virtual Audio Cable) makes a fake audio device, used to re-route the audio of different programs

In Wokada context, it's used to get the output of wokada as the input in other programs

Also there's no working wokada colab for free

And next time pls elaborate in #1192011222023950368

[covert lake]

Elaborate ur specs in #1192011222023950368

[chilly lake]

there are better audio enhancer models

I think this one https://github.com/yxlu-0102/AP-BWE

I dont think so, you need a local gpu

technically you can probably run it on colab as well

yeah

same as usual - clone repo, make virtual environment, activate it, install requrements

[glad nebula]

ap-bwe uses a ton of vram

[covert lake]

Just buy an RX 9090

[glad nebula]

[covert lake]

No, just the online port

https://github.com/Nick088Official/versatile_audio_super_resolution

[atomic pilot]

hi guys

[chilly lake]

numpy 2.0+ got installed

instead of 1.26.4

in colab !pip install numpy==1.26.4 in a new cell

[wanton citrus]

Is the Deiteris' W Okada Fork really the best live voice changer in the market?

[chilly lake]

hm... then it had okay numpy before

1.23.5

maybe

anyway, just restart the cell

https://www.python.org/ftp/python/3.11.9/python-3.11.9-amd64.exe

[mild delta]

heyyy

[chilly lake]

open .cmd exe and type where python

[chilly lake]

you unzip the file and move these into checkpoints folder

after that you use the path to the model you need

based on the input file sr

[tender siren]

how can i make my own i18n? im makin applio plugin and i wonder, does it load automatically from .jsons from certain directory?..

[chilly lake]

i think it tries to use the main applio i18 files by default

if you're trying to upscale 24k audio to 48k, you provide a path to 24to48

python inference_16k.py --checkpoint_file ../checkpoints/24kto48k/g_24kto48k --output_dir ../generated_files/24kto48k

[tender siren]

# Initialize i18n
from assets.i18n.i18n import I18nAuto

i18n = I18nAuto()

i guess it doesnt take any arguments does it

is there non-auto lol

[chilly lake]

I18nAuto is imported in main app.py

so it automatically picks the files from assets/i18/lanuages based on the configured language

if you want to make your own, make languages folder in your plugin and init your own version of i18n

24k is 24k

[tender siren]

im f-ing genius

[left nova]

guys

why can't i hear anythin when i use okwada vc

[tender siren]

Own languages in your addon

Tested on 1 but will probably work for all as Applio requires restart on language change

# `plugin.py`
import os
import gradio as gr
from assets.i18n.i18n import I18nAuto

INITIAL_LANGUAGE_PATH = I18nAuto.LANGUAGE_PATH
try:
    I18nAuto.LANGUAGE_PATH = os.path.join(os.path.dirname(__file__), "locale")
    i18n = I18nAuto()
except Exception as Error:
    gr.Info(Error)
finally:
    I18nAuto.LANGUAGE_PATH = INITIAL_LANGUAGE_PATH
    i18n = I18nAuto()

Where your file structure is:

📁 my_plugin/
├───📁 locale/
│   ├───📄 en_US.json
│   └───📄 pl_PL.json
└───📄 plugin.py

~~⚠️ You will either have to translate your plugin to all languages from .../assets/i18n/languages directory or never change language to one in which plugin is not translated in. It will result in:

  File "X:\ApplioVX.X.X\assets\i18n\i18n.py", line 40, in _load_language_list
    raise FileNotFoundError(
FileNotFoundError: Failed to load language file for xx_XX. Check if the correct .json file exists.
```~~

[chilly lake]

it may work

it loads the mapping at the start for the main process

so this way you do create a second instance with new path

you need to make en_US.json at least

[tender siren]

i think its ok now

:d

[bleak vector]

that spectrogram depicts a.. most likely 32000hz audio that is so heavily compressed to the point it's spectrum response maxes mostly around 16k. Let's just say it originally was a 32khz [ ignoring the compression ] audio file saved as 44khz - maybe upsampled by an accident?

well.. just a thing to always remember when reading specs is that you want to double the frequency range
edit: So in your example, you see ' 22 khz ' on top, that would indicate 44.1khz ( 44100 hz )

what do you mean.. ?

ah, I wouldn't say weird but simply ' lossy '

[lusty dome]

greetings fellow ai users

i am buttcheeckman

ruler of BBC

[bleak vector]

oh, and if you did not know.. a lot of things can cause the audio, and effectively spectrograms to become ' lossy '
certain codecs such as mp3, opus or vorbis are great examples.

Yes but, try to recall what it was before you got the wave. What was the source? Perhaps you recall the extension it had?

Oooo then yes, that is the culprit

mp3 from all of those I mentioned is the worst..

[potent helm]

which models best for turkish do u guys know

[tepid rapids]

ok so will gtx 1650 sup + ryzen 3 3100 will be good on w-Okada cause i hear a lot of background even if i put steelseries gg mic or when i talk on my native language i don't think it's trained on my language cause like there is some word that ai can't say it and it will be obvious to anyone that it's ai

[hazy swallow]

hii

[finite grove]

Hey, I am a ai automation developer if you need any help or service just drop me a msg @finite grove

[covert lake]

#1192011222023950368

[bold rampart]

Bruh, I just found a new, open-source AI video generator, it's here: https://hunyuan.tencent.com

[polar flax]

converting to another lossy format worsens the fidelity

the best practice is to always convert to 32-bit float wav during processing

perhaps you mean the lossy compression that not only cuts off frequencies but also causes amount of hollows depending on the bitrate and lossy format

[chilly lake]

there's an implementation called FramePack

and a specific implementation of it in SD.Next that can do unlimited length generation

[bold rampart]

Yeah, I already tried installing locally before, but it was really slow, and I was becoming impatient

I ran that on my laptop which has a 12th Gen Intel CORE i7 CPU, 16GB RAM, a NVIDIA GeForce RTX 3050 Ti Laptop GPU with only 4GB of dedicated VRAM

[chilly lake]

16GB VRAM, 48GB RAM can cook it

[faint smelt]

alguém tem o link do site de IA cover?

[bold rampart]

I said 16GB of RAM, not VRAM, my real VRAM is only 4GB

[naive nimbus]

could anybody help me

it wont work for me

[chilly lake]

and I said what you really need

[bold rampart]

Yeah, I don't have that

[naive nimbus]

when you press start are you supposed to hear your own voise

[hollow pollen]

so i have a question for someone that could help me. so i have a real time voice llm but i would like to make it so i can have it in a discord server so me and my friends can have fun with it how would i do that? link to anything would be so helpful

[quiet oriole]

is lisp good for ai

[faint smelt]

Does anyone have the link to the collab? The only one I had is no longer available

:/

[echo inlet]

isn't this the one used way back? like for ancient ai lol. people use python nowadays for it

[echo inlet]

BERTurk as it is trained in turkish text

[polar flax]

(+ (* a b) (* c d))

[vague herald]

Anyone know a site I can pay to train an AI vocal model

[barren merlin]

anyone know how do i download ai model zip file on weights

[wraith geyser]

hello! can i ask where can we train voice models with a pre-train?

i really wanna try using different pre-trains, but apparently only few are available on applio

[glass otter]

hi

how to create a image with AI ?

[vague herald]

Anyone know a site I can pay to train an AI vocal model I have my data base just need someone to run it to 1000

[slender moss]

whats the best ai video generating tool free ofc

[potent helm]

can u send me the model

or which girl voice model best for turkish

[still sigil]

@sharp oracle no advertising

[dense badger]

hi

[bold rampart]

I made ChatGPT VERY rude...

[potent helm]

There is a Roblox YouTuber named Kira. How can I make my voice sound like in her videos? Can anyone help me troll?