🧬│ai-chat

[fallen kettle]

ok wheres the online one?

[covert lake]

As you dont got a good PC, its better you use cloud for training an RVC Voice Model:

• Google Colabs (4 hours of daily gpu for free, not much hours, but easy to use):
• Applio (ui)
• Mainline (UI)
• RVCDISCONNECTED (no ui)
• Kaggles (a bit harder to use and needs phone number but gives 30 hours weekly of better gpus):
• Mainline (UI)

Yw

[little sky]

do i have to redownload the program or just adjust the settings

[polar flax]

use cpu mode in the fork wokada if possible https://rentry.co/ForkVoiceChangerGuide

[ionic pumice]

is your grandpa playing in a cod tournament

[little sky]

is this safe ?

[covert lake]

Yes dw, it's made by staff

[little sky]

okay ill try it

[polar flax]

having just a good 144-240hz OLED monitor is enough, even mid end gpus can already push fps more than that refresh rate

[polar flax]

don't trust avast's false positive or lookalikes

[little sky]

uh what do i do now? im in a website

[feral halo]

Ayo? @little sky level 2 !!!

[native mauve]

e

[obsidian quiver]

wassup

[little sky]

i got it working lets goo

[zealous zephyr]

ay guys how can I rephrase words from a movie dialouge and make the characters say something funny instead of their actual dialouge

[ionic pumice]

funny number (90% are bots)

[glass trellis]

how to chat in #✦│chat

?

[elder willow]

I have good news, I got my best friend back and now I'm happy. 🥰

[covert lake]

u need to get level 5 by chatting

[glass trellis]

much much time it needed to be in level 5 @covert lake

[covert lake]

you'd have to say it urself and then use the model over ur voice, or either use TTS, like Applio and Ilaria RVC got tts but rvc is not really the best for tts if u want emotions

promo only in #1159290752195633273

[covert lake]

u just need to chat

[fallen ruin]

https://tenor.com/view/laguna-ju-ju-pixela-ju-ju-pix-ju-ju-gif-23004562

[ionic crystal]

Even igpu is better

[fallen ruin]

https://youtu.be/O9mmp8SIrBQ?feature=shared https://youtu.be/_EurWOTJpfo?feature=shared

LMAO

I THOUGHT MARKUS "NOTCH" PERSSON PASSED AWAY

[covert lake]

promo only in #1159290752195633273 , everywhere else ur promo will get deleted

[late pier]

hi

[covert marlin]

всем ку

я с однйо фигнёй разобратся не могу

хееелп

[cyan geyser]

what is epochs

что такое epochs

[covert lake]

https://docs.aihub.wtf/rvc/resources/epochs--tensorboard/#epochs

[covert lake]

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

[covert lake]

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

[covert marlin]

?

[stable furnace]

Kaka

[pine saddle]

Hi, is Applio Training working on Colab? Or do you still get disconnected after some minutes?

[covert lake]

it should work all fine, however i would suggest to use Kaggle as it gives more hours

[pine saddle]

Is it free?

[feral halo]

Ayo? @pine saddle level 1 !!!

[ionic pumice]

didnt u make those old rvc colab tutorials .,,,

[half wing]

I'd use kaggle if I understood

[covert lake]

Yes

[half wing]

But I'm on mobile and I can't be bothered

[covert lake]

As you dont got a good PC, its better you use cloud for training an RVC Voice Model:

• Google Colabs (4 hours of daily gpu for free, not much hours, but easy to use):
• Applio (ui)
• Mainline (UI)
• RVCDISCONNECTED (no ui)
• Kaggles (a bit harder to use and needs phone number but gives 30 hours weekly of better gpus):
• Mainline (UI)
• Applio by Vidal (UI)
• Applio by Shirou (UI, no guide though)

[zealous zephyr]

oh

alr thanks man

[half wing]

Rvc disconnected is great for training

[feral halo]

Ayo? @half wing level 3 !!!

[pine saddle]

Haha yes, I’m making a new updated one here’s why I’m asking x)

[covert lake]

i can tell you how to do it with tts if u want, just it doesn't have emotions

[half wing]

I also used applio no ui

[pine saddle]

Thank you 🙂

[covert lake]

kaggle offers more free gpu time so thats why is suggested more, even if its harder tho

[half wing]

You're the dude who saved my life when I started

[zealous zephyr]

yeah what's tts?

[pine saddle]

❤️

[half wing]

I have like 5 Google accounts so I'm okay with 4hours of GPU in a day

[half wing]

I'm still learning how to continue training when it stops tho

[covert lake]

Text To Speech,
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

[pine saddle]

Why do you consider it harder?

[half wing]

Ik I can since I use my drive but I have never needed to

[covert lake]

Maybe bark tts from the index + rvc has more emotions, but less quality

[zealous zephyr]

oh my bad

thx man

[half wing]

How do you guys resume training?

[ionic crystal]

@covert lake did you know training a model is a ssd health killer

[half wing]

On collab

[covert lake]

wait what

@pine saddle what tut did u make?

[half wing]

Onions is the og tutorial

[ionic crystal]

I had like 100gb in my logs foldee

[zealous zephyr]

applio don't have tts right?

[half wing]

For easy gui

[covert lake]

the one of spongebob rvc ?

[half wing]

Yep

[covert lake]

was it a yt tut

[half wing]

Yes

[covert lake]

U made this?? https://youtu.be/hSxTLCR_95Y

LMFAO

[half wing]

I've been making models for over a year and nevery tried to get model maker here so I'm attempting 💀

[covert lake]

I REMEMBER SEEING THAT THE FIRST TIME I STARTED ABT RVC

[half wing]

Not me, onions

[covert lake]

it requires the phone number, and grok, and the ui is a bit harder as it shows the code for an average user

[covert lake]

yea i pinged the wrong guy

[covert lake]

huh

[pine saddle]

I did haha

[covert lake]

idek whos that

[half wing]

Who gives model maker? I submitted but I wonder

[pine saddle]

Ahh got it

[covert lake]

lmfao didn't expect u'd be still there

[zealous zephyr]

@covert lake does Applio have tts?

[covert lake]

Yes, its wrote in the message i sent that they use microsoft edge tts api for tts

[zealous zephyr]

ohh I see

[half wing]

Nickk

[pine saddle]

Yeah Google decided to entirely shutdown the easyGUI Colab so I’m forced to make an update video

[zealous zephyr]

ayo are you that raw onion dude on yt?

[pine saddle]

yes x)

[covert lake]

easygui colab was broken even before shutdown #📰│dev-updates message , the creator left developing since months

[half wing]

Everyone started with you I'm pretty sure

[covert lake]

QCs (Quality Controllers)

[zealous zephyr]

ayy man your videos are awesome so many people recommened your videos when I started using AI

[half wing]

Okay 👍

[covert lake]

on what are u gonna make the next video? like there are some different colabs and kaggles and now people use pretrains

[half wing]

Someone told me titan was the worse but idk why

For singing that is, since I only make singing models

[covert lake]

yea, its the hardest way but also the best way, 30 hours weekly of better gpu so even faster

[covert lake]

some people dislike pretrains, some don't, pretrains aren't 'rvc v3', they are mostly helpful for other languagues though

i made a pretrain guide

https://docs.google.com/document/d/1j9J8A8Oop9bMOHmCs3jDXzPujuD6TQ0Q396rJ0MyuIc/edit

[pine saddle]

I thought it worked (at least for inference), I remember I tried it some time ago

[pine saddle]

thank you 😁😁

[covert lake]

nah it couldn't even download the program sucessfully so coudln't even run the ui, and i seen the code myself and the no ui inference had some outdated commands which didn't work

it had a problem with pip dependencies

[zealous zephyr]

and btw how can I create AI interrupting vids man that website from which you did in your latest video is asking for money now

[half wing]

Ai what

[covert lake]

for inference most people use ilaria rvc zero which is the fastest for cloud as its a zerogpu huggingface (A100)
While for training it kinda depends, the best one in terms of performance is one of the kaggles, i shared u 2 kaggle guides (mainline & applio)

[covert lake]

u mean those tiktok memes where time traveler disrupts an old meme?

[zealous zephyr]

yeah

[pine saddle]

I’m making a whole video for every type of user, I’ll show how to use weights(dot)gg for mobile and casual users since it’s free for inference and very intuitive, then local inference/training with Replay for those who have a good GPU, and at the end I planned to show how to use RVC Disconnected, but now that you’ve shown me Kaggle I’m wondering which one is better

[pine saddle]

Ohh I see

[covert lake]

personally, ig only luma labs, they should give only 30 credits monthly for free, u should be able to see when they refresh in https://lumalabs.ai/dream-machine/account

[pine saddle]

oh really? wtf

[pine saddle]

Yeah that’s what I knew too

[feral halo]

Ayo? @pine saddle level 2 !!!

[zealous zephyr]

bro I logged from other id

[random fulcrum]

Hey yall, im doing a sort of project and needs yalls help.
Do any of yall know give or take around 5 llms with a decent API such that I can have a way of coding in python automating messages between em? Also, with the ability to package some parts of the response as some metrics as part of a seperate json file? They dont need to be running concurrently, they can be ran at separate instances and dont need to remember any of chat history, each API call to each separate llm only needs to know a set prompt.

[zealous zephyr]

oh wait I

[covert lake]

oh idk much of local, but usually people use Mainline or Applio locally afaik

I used RVC Disconnected too, but Kaggle is just better in performance than google colab, as u don't have to worry about the training stopping bc of finished gpu as u have 30 hours weekly of more powerful gpus, so thats why its better

[zealous zephyr]

I logged from another account and it is working les go

[covert lake]

yea its 30 free monthly credits per account

[pine saddle]

Can you also stay AFK with Kaggle?

[covert lake]

about inference on cloud, usually people either use Weights.gg, Ilaria RVC Zero or astra labs
(I personally used ilaria rvc zero)

[zealous zephyr]

[feral halo]

Ayo? @zealous zephyr level 3 !!!

[covert lake]

Ngl, i dunno that, I haven''t heard of people getting disconnected for being afk, but i also remember starting a session for other things, forgot i started it and went doing other things and after it had like a popup 'are u still there?' but i don't know if that happens only when u stay afk not running anything or also when u run something

@molten stone u know anything about being afk in kaggle?

[molten stone]

sometimes its disconnects

but most of the time it doesnt

[covert lake]

so it happens even while running the cell?

[molten stone]

but from what ive seen it only disconnects when theres nothing happening in the code, no output

[molten stone]

if theres no difference in output or new line basically

it disconnects after a looong time

idk tho since that only happen to me a few times before

[covert lake]

so not while training as the output changes right?

ig this should work anyways https://docs.google.com/document/d/1Pr-AZndodmWgsbOeuHQU4IrgbatFgYc1ChOq_ZAf_5s/edit#heading=h.sgz3srgioeqz

[molten stone]

ye it shouldnt

but not quite sure since it only happened to me on those specific situations

and i mostly run on headless

[covert lake]

may be dumb question but wdym headless

[molten stone]

running using the save & run all button

[molten stone]

what they said XD

[covert lake]

So what i found out is:

• Maximum runtime session is 12 hours no matter what

• The interactive editor isn't supposed to pop the "are you still there" modal unless it detects no edits, or cells not running for 40m/20m

• Save & Run All will run the notebook in a totally separate VM in the background, separate from the 'interactive, editable notebook', Save version> Run all (confirmation), and will expire only if it is performed for> 9 hours (or 3 hours for TPUs), also you can close the tab and come back later to check on the version result (it won't show in the interactive editor, you'll need to look at the version created by the commit to see results).

I didn't webscrap it i just read all comments 😭

i have only heard of https://github.com/ScrapeGraphAI/Scrapegraph-ai

[molten stone]

yup thats the reason i wanted to add file manager so that even running on headless you can still access the files

[feral halo]

Ayo? @warm loom level 2 !!!

[covert lake]

this server isn't about webscraping, its mostly about RVC & Wokada

i don't really need webscraping

[molten stone]

also on headless, its still 12 hours even if nothing is happening, and you are given an additional 9 hours at the end of your 30 hours IF you run it with something like 4 hours left in your time

[covert lake]

inifinite gpu glitch?

[molten stone]

na only 9 hours

[covert lake]

still 9 hours u shouln't be able to get tho

🤑

[molten stone]

its probably given as a "just in case" kind of thing

[covert lake]

did kaggle literally tell u ur gpu eneded but for headless they will continue or u just found out randomly

[molten stone]

very useful when training multiple model at the same time

[half wing]

How does one become verified

[molten stone]

just found out randomly

[covert lake]

by chatting till u get level 5

[molten stone]

and i timed it

multiple times XD

[half wing]

Smh

I don't have that many things to say

[covert lake]

sushi or pizza

[covert lake]

lol

[half wing]

Pizza

I'm praying my model gets accepted

[covert lake]

agreed

[half wing]

It's really clean

I'd sent a snippet but I am not allowed

Sobs

[covert lake]

btw, does kaggle have a captcha for afk ever like colab?

like for new accs or sometimes idk

[molten stone]

never got it so idk?

[covert lake]

thats good

[molten stone]

or never noticed

[covert lake]

yea was just asking in case i was thinking writing about a 'Kaggle Workarounds' like a did for colab

[half wing]

Maybe one day kaggle will work on mobile or idk

I might get a computer but mobile training kinda slays

[molten stone]

dont think theres any need for workarounds on kaggle

since theres nothing to do workarounds on

[covert lake]

i never tested kaggle on mobile but it should work, just the interface is not really mobile-friendly

[half wing]

I went on kaggle once and recoiled lol

But I'll warm up if collab bothers me

[covert lake]

i guess writing info about the AFK thing and about watching out if they using an old kaggle which has disallowed code so automatically turns off the session

[molten stone]

i forgot which are banned

i think only facefusion?

but idk if its been lifted or not

[half wing]

I'm lost

[elder willow]

Hello guys I need help making content trough ai

[molten stone]

been running my voice changer normally on kaggle and its not disconnected yet

been 2 hours

[elder willow]

Guys?

[covert lake]

Okay so everything technically banned on kaggle is:

• RVC (just stops session)
• colabcode (just stops session)
• Facefusion (insta ban)

[half wing]

Why I RVC banned everywhere

[feral halo]

Ayo? @half wing level 4 !!!

[covert lake]

do you need to make ai covers?

[half wing]

Finally level 4

[covert lake]

because alot of people will abuse cloud services only for rvc

[half wing]

So how are we using this for rvc

[elder willow]

More like text remover and language transcript so I replace the already existing video with other language

[feral halo]

Ayo? @elder willow level 1 !!!

[covert lake]

so like remove subtitles and add subtitles of other languages? dunno about that sorry

[molten stone]

rvc seems to run just fine on normal but on headless theres a chance it will stop at installation

but onces its done with installation it will keep working no problem

[covert lake]

rvc without encryption instantly shutdowns session

[molten stone]

weird

[covert lake]

im talking about no encryption, ofc u can bypass the things i said

but yes

i rememeber tying to make an rvc mainline kaggle

much much time ago (before ur notebook)

and it instantly shut down without encryption

[elder willow]

i had kaggle cut me off when i used mainline xd

[elder willow]

Know someone that knows so I can message?

[earnest dragon]

Finally only works on level 5

[half wing]

Ik sobs

I'm trying to yap but idk what about

And I'm not about to spam lol

[covert lake]

https://www.reddit.com/r/StableDiffusion/comments/16ehr4x/welp_im_done_with_kaggle/ i think it insta bans when using automatic111 or comfyui?

Not the diffusion library tho, i runned Flux.1-dev in kaggle for inference

[half wing]

What's the difference with normal and local?

[covert lake]

ye without encryption it gets detected

[covert lake]

dunno sorry

[covert lake]

normal? u mean cloud/online ?

[half wing]

Yeah

[molten stone]

im runinng comfyui

wot

[covert lake]

without any encryptions?

[molten stone]

i don remember

[elder willow]

I have a banned account in kaggle for making a DeepFakelab port.

[molten stone]

lemme chck

[covert lake]

i got banned for facefusion

[elder willow]

lol

[covert lake]

google hates very much deepfakes

[elder willow]

good thing I have my family's phone numbers to verify the multis xd

[half wing]

So what is the difference 😓

[covert lake]

i heard rope unleashed is not banned tho

[earnest dragon]

Yeah I know a program that can do this that is open-source

[covert lake]

Cloud/online: runs the code in a remote good pc
Local: u run the program on ur GOOD pc

[half wing]

Yeah so don't do it on phone? Lol

[covert lake]

#1159290752195633273

[covert lake]

u can use cloud on phone

[twilit furnace]

h

[elder willow]

fixed some little things in uvr5 ui kaggle as like it installed 2 times ngrok xdhttps://www.kaggle.com/code/ahaghjhj/uvr5-ui-kaggle

[half wing]

I use cloud already

I meant local

[elder willow]

Which is it please

[ionic crystal]

Voice ai is so dead

[earnest dragon]

Yeah

[covert lake]

nah, phones can't train, MAYBE if u install termux u COULD inference (use model) on CPU, but i dont think anyone ever did that

its just better to use cloud

[half wing]

Yeahh

[covert lake]

[elder willow]

no longer generate views to make ai covers

[covert lake]

i remember when a famous italian streamer listened to one of my ai cover in live lol

[half wing]

I do ai for kpop so it's pretty alive

[covert lake]

now ai covers hype is kinda ded

[elder willow]

mobiles on the grill 😜

[half wing]

Everyone wants to see their fav group sing other group songs

[elder willow]

@earnest dragon ?

[ionic crystal]

Kpop is also dead

[half wing]

It's really not lol

[ionic crystal]

It is.

[covert lake]

lmk cus its interesting

[ionic crystal]

Just because only you do it it doesnt mean its alive

[half wing]

A lot of people post everyday?? And I get a good thousand of views depending on the video??

[ionic crystal]

No

[covert lake]

look, ai covers are kinda ded, idk about kpop but don't argue yall srsly

[half wing]

I'm not arguing I'm just like ??

[ionic crystal]

@covert lake

[covert lake]

was just saying

[covert lake]

perfect

[molten stone]

no-

last run 3 months ago

so idk

[ionic crystal]

From ipython import javascript

[elder willow]

how do you start imjoy elfinder, i tried to implement it in my uvr5 ui in kaggle and i got desperate with the ports

[covert lake]

btw do u mean deepfacelab ?

[elder willow]

yes

[covert lake]

LMFAO

[molten stone]

em using ngrok

[covert lake]

I'd try using automatic111

[molten stone]

the commented one is for imjoy

[molten stone]

ban

[ionic crystal]

@molten stone

[elder willow]

I use ngrok but apparently imjoy's localhost does not start

[molten stone]

just copy paste mine

[half wing]

Cursed

I hope it's so bad that it's good

[earnest dragon]

https://pyvideotrans.com/

[earnest dragon]

yeah kaggle doesn't like deepfake stuff

[elder willow]

kaggle is from google lol

[molten stone]

use mine its not even my code XD

[molten stone]

ye but different team

[elder willow]

that's what I did xd, part of it is mainline code

[earnest dragon]

Yeah and also Google colab doesn't like it either

[covert lake]

could u share the whole notebook if u can please?

[molten stone]

XD

[molten stone]

dm

[fallen ruin]

Michael Jack meets the touhou girls

[dreamy heath]

hi

[elder willow]

lol

works

I'm going to mute imjoy's output, I don't need it xd

[elder willow]

works lol

[molten stone]

lol

[molten stone]

true

i jsut keep forgetting

[elder willow]

use a modified to avoid problems with gradio lol

https://github.com/Vidalnt/imjoy-elfinder

[molten stone]

never got any wot

the only problem im getting is when downloading it takes more bandwitdh thatn it should

[elder willow]

current version of uvr gradio has problems with imjoy

[marsh jasper]

hi

[feral halo]

Ayo? @marsh jasper level 1 !!!

[marsh jasper]

hi

is anyone alive

[deft oak]

what if i trained voice in elevenlabs and then use elevenlabs audio to train again in rvc?

[elder willow]

https://tenor.com/view/lorry-accident-bridge-crash-gif-13667303

[marsh jasper]

which course is best for begginers for mle?

[radiant vector]

Que onda? alguno tiene el voice model de Momo y coscu?

[ionic crystal]

Google

[ionic crystal]

Muchas gracias afición, esto es para vosotros Siuuu

[wild locust]

hi everyone

[ionic crystal]

Hello weirdo

[marsh jasper]

really

[elder willow]

https://applio.org/models

[marsh jasper]

how google

[fallen ruin]

🐠

[radiant vector]

sos un crack

[elder willow]

yonofumocrack

[radiant vector]

que

[elder willow]

so

[stiff heart]

Yes my bad brother

[gentle canopy]

hi

[feral halo]

Ayo? @gentle canopy level 11 !!!

[eager finch]

hi

[cyan temple]

Hi everyone

[drifting crag]

hi

[spark jetty]

hi

[half wing]

I think RVC disconnected trains the fastest but idk

I'm trying to figure out which one takes less time

[rose laurel]

hi

[fallen ruin]

https://tenor.com/view/bondrewd-made-in-abyss-cool-gif-26781094

[elder willow]

i ported rvc studio to kaggle lol

-kaggle

[rare sorrel]

📘 Kaggle Notebooks

• Applio Notebook, by Vidal Kaggle
• Applio Notebook, by Shirou Kaggle
• Music Source Separation, by Shirou Kaggle
• UVR5 NO UI, by Eddy Kaggle
• RVC Mainline, by Hina Kaggle
• Original W-Okada's Voice Changer, Kaggle
• Modified W-Okada's Voice Changer, Kaggle
• 📖 How to use RVC Mainline Kaggle by Cauthess

Note: Kaggle limits GPU usage to 30 hours per week.

[elder willow]

Interesting UI, is that MAC or Linux? Looks cool

[elder willow]

kaggle

[elder willow]

Thank you

[twilit furnace]

uh why my F0 det :" only have rmvpe_onnx and crepe_tiry"

[feral halo]

Ayo? @twilit furnace level 1 !!!

[twilit furnace]

i dont have like dio/crepe.....

and i cant del the model

[earnest dragon]

amd or nvidia

[twilit furnace]

nvidia

[craggy owl]

Hello all.
An AI cover I really love has been removed from YouTube. It was a GLaDOS cover of 'Little Dark Age'. Does anyone have it?

[stuck creek]

best mvsep main vocal isolation model? are the aihub docs up to date (UVR-BVE-4B_SN-44100-1)?

[feral halo]

Ayo? @stuck creek level 3 !!!

[twilit furnace]

it different for the fils direcML and gpu?

[elder willow]

https://www.kaggle.com/code/ahaghjhj/rvc-ai-cover-maker-ui/notebook

[craggy owl]

Maybe the creator of the piece is even on here?

[feral halo]

Ayo? @craggy owl level 1 !!!

[elder willow]

has some bugs xd

set 1-2

works lol

ok, i will fix it

[turbid pivot]

fuccckkk man, keep on posting memes like that man

god dam fuckhead

[sleek elk]

hello

[tawny oracle]

hi, any working zerogpu ai cover generator?

[elder willow]

-hf

[rare sorrel]

Suggestions for @tawny oracle

<:huggingface:1179800228946268270> Hugginface Spaces

• UVR5 UI, by Eddy and Ilaria Huggingface Spaces
• Ilaria RVC Zero, by thestingerx Huggingface Spaces
• RVC⚡ZERO, by r3gm Huggingface Spaces
• Applio, by IA Hispano Huggingface Spaces

More commands: -audio, - colabs, -kaggle, -local, -overtraining, -realtime, -rvc, -help

[tawny oracle]

thx

[white sky]

I want to use a human voice in a video, but in the Generator voice style, but the voice is realistic and not robot-like (a voice that can have voice tone and settings). How can I do it or where can I find it?

[mellow echo]

herllo

[lavish zenith]

hi

[covert lake]

ay that's good @sturdy elbow

[elder willow]

Yer

[elder willow]

i want fix some bugs with the gpu lol

[steady basalt]

hello

[pine saddle]

@covert lake Did OG RVC have a pretrained model by default on?

[steady basalt]

i wanna ask yall sum. is it possible to detect if a picture is ai made for free?

[pine saddle]

(with OG RVC I mean the first Colabs, including EasyGUI and Kalomaze’s)

[covert lake]

yes, since rvc exists rvc always had a pretrain from default, its just now colabs have options to change it as now there are new community pretrains

[pine saddle]

👍 Thanks

[covert lake]

but yea, every model since rvc ever developed had a pretrain on default, without it, it would take an insane amount of hours (called training by scratch), while pretrain help to prevent that

Yw

the original pretrain is trained on an english-only dataset

[dry fox]

Hey anyone have a suggestion for a deep voice?

[dreamy lava]

xfnsfsfn

[modest flint]

hi

[fast pilot]

hi guys

[dusk sable]

English or Spanish?

[sturdy elbow]

@elder willow @covert lake Yeah, I made some parts of the code with kaggle in mind, I just haven't had time to test it, I'm a bit busy with applio

[earnest dragon]

Yeah applio I prefer over all the other rvc ones

[vale patrol]

Hi, human exploring this world

[earnest dragon]

Don't talk about iptv here and I already know the best one avaliable that really has no competition.

[elder willow]

HEY CAN I PARTNER

I would love to partner with this group

[covert lake]

alr

[covert lake]

piracy 🔥

[earnest dragon]

Yeah but he is probably promoting a trash one when there is only one good one that allows you to request channels to then they will add them

[fallen ruin]

https://tenor.com/view/kyojuro-rengoku-anime-gif-21337946

[unreal moth]

hello, wondering whats the best ai live voice. theres to many im not sure which one to pick

[feral halo]

Ayo? @unreal moth level 1 !!!

[unreal moth]

or whhat one to pick. i used to troll my friends with this one ai voice but it was years ago, and now i have no idea what one i should download

if sombody could tell me it would be greatly appreciated

real time voice changer"

before im pretty sure i used rvc i think

[white notch]

Hi

Should I use RVC if my graphic card is an AMD RX 570?

[storm meadow]

-rt

[rare sorrel]

This interaction has expired, use the command /guides realtime if you wish to see it again.

[elder willow]

sup

[sullen sentinel]

[feral halo]

Ayo? @sullen sentinel level 1 !!!

[azure nova]

Hello World, greetings from Mexico. I am interested in the deepfake system for both webcam and text to speech in Spanish language. I would like to belong to your community.

[earnest dragon]

This server isn't for webcam but my discord is for live deepfakes and stuff but not one image.

[high patrol]

Hello here 🤗

[golden shard]

How can I create a java application for a pdf summariser?

[feral halo]

Ayo? @high patrol level 1 !!!

[high patrol]

Who is an content creator here

[fringe haven]

What's the most CONVINCING girl voice changer

[elder willow]

i will make it for u

[void pulsar]

Hi guys! Does anyone by chance know if there is Kamala Harris voice model out there? I've been assigned to do a political presentation and it would be a great help! Thanks!

[ionic crystal]

No.

[elder willow]

if you don't know how to program use this https://www.blackbox.ai (it's not the best compared to knowing how to program but if you drive it to what you want to achieve it works)

best option: use brain

[ionic crystal]

If you drive it???

I think you mean if you run it

[elder willow]

https://youtu.be/xk4_1vDrzzo?si=0P-bNFiV6lbwtPcA

[clear summit]

Question

[ionic crystal]

Answer

[clear summit]

Ima be honest, i just joined, and i've been using weights since june

but where can i use rvc models

[elder willow]

-colab

[rare sorrel]

Suggestions for @clear summit

☁️ Google Colabs

• Applio, by IA Hispano Google Colab
• RVC Disconnected, by Kit Lemonfoot Google Colab
• RVC Mainline, by Hina Google Colab
• AICoverGen-WebUI, by Hina Google Colab
• AICoverGen-NoWebUI [English], by Ardha, fixed by Eddy, Hina and Gdr Google Colab
• AICoverGen-NoWebUI [Spanish], by Eddy, Hina and Gdr Google Colab
• UVR5 NO UI, by Eddy Google Colab
• UVR5 UI, by Eddy Google Colab
• Modified W-Okada's Voice Changer, Google Colab

ℹ️ Note

While the Colab free plan provides up to 12 hours of daily usage, the GPU is typically available for only about 4 hours each day on average.

More commands: -audio, -hf, -local, -kaggle, -overtraining, -realtime, -rvc, -help

[elder willow]

choose

[clear summit]

Uh

The last one i guess

[ionic crystal]

@elder willow @clear summit

[elder willow]

the last one is for realtime

[earnest dragon]

Nah Cursor ai is better

[clear summit]

O

[feral halo]

Ayo? @clear summit level 1 !!!

[elder willow]

the brain is better

[elder willow]

use applio

[clear summit]

Alr

[elder willow]

[earnest dragon]

Yeah but the brain and Cursor ai is better because then it saves sometime with certain stuff.

[clear summit]

I clicked on the link beside the thing

now whar

[elder willow]

click "install applio"

[clear summit]

The dropdown?

[golden shard]

Thanks for this
And I have one more question like how can I create an AI model in java for the pdf summariser

[elder willow]

here

[elder willow]

idk

[golden shard]

Ok

[elder willow]

use kaggle?

[clear summit]

Oh wait

so i run it?

[elder willow]

yes

[golden shard]

Yeah but creating of using an AI model in java is tough I need to learn that first

[feral halo]

Ayo? @golden shard level 1 !!!

[elder willow]

I like to use python better xd

[clear summit]

It says executing in num. seconds at the bottom. Am i doing it right?

[golden shard]

Same but the project is in java because of different operating system

[gloomy vector]

Hi

[elder willow]

mmm

pass screenshot

[clear summit]

No perms

It completed just now

[elder willow]

now run "Start Applio"

[clear summit]

click the play button?

[elder willow]

yes

[clear summit]

It just opened something called tensorboard

[elder willow]

ignore it

[clear summit]

Now it's downloading

[elder willow]

is for making ai voice models

[clear summit]

Do i wait for it to stop loading?

Also it gave me two links

[elder willow]

click on the link named "Public URL"

[clear summit]

Alr, now what

[feral halo]

Ayo? @clear summit level 2 !!!

[elder willow]

search a model in #1175430844685484042

[clear summit]

I already have a rvc pth thingy downloaded

[elder willow]

read this https://docs.applio.org/getting-started/download-models

[clear summit]

I see. Do you need an index file?

[elder willow]

yes

[clear summit]

Mmmk

[elder willow]

rvc uses pth and index

[clear summit]

Alr

[clear summit]

Wait how do i upload the pth file? Sorry i'm asking alot, i'm a lil braindead rn

[polar flax]

python is easier to learn, but if you want strongly typed programming, learning java or C# is worth it, though C# is more preferrable for game development

[elder willow]

in "download" tab

[clear summit]

Yeah

[polar flax]

either using the "download" tab, or doing manually by putting files in the logs folder like following, and then click Refresh in the inference tab

[elder willow]

To download a voice model, go to the “download” section and enter the link to the file. Now you can search for models from Applio by simply entering the character name to search in the "Search Model" section, then copy the model link or download it.

[elder willow]

your version of applio is old

[polar flax]

I also have the latest 3.2.x, but that old one is quite stable

[clear summit]

I uploaded both files, now what

[elder willow]

make ai cover

[elder willow]

Making a single Inference
Step 1: Select your model
Simply select in the boxes the .pth and .index file of the downloaded/trained model. If not displayed refresh the model list.

Step 2: Load your audio
As the first step, select and upload the audio file in the corresponding box.

Step 3: Make the inference! 🎉
now you are ready to perform your inference, just click convert and you will get your result. The processing time will depend on the speed of your CPU/GPU.

[clear summit]

Oh alr bet

[elder willow]

that version makes very slow inferences compared to current versions

[clear summit]

You can't record audio there?

[elder willow]

yes

[clear summit]

Aw. Show.

[clear summit]

Aw. Show. 2x

[elder willow]

best click on "upload an archive" xd

[wide cargo]

I need my ass some nitro desperately

[mystic frost]

Hi

[wide cargo]

Wsp

[trim osprey]

LOCK IN

imma make AI make me sound like a pretty girl

[mystic frost]

Where is the chat in Spanish?

[mystic frost]

(I support myself with the translator)

[wide cargo]

--okada

FUCK

[trim osprey]

hi i'm smooth brain where do i find the google sheet with all the voice models again?

[mystic frost]

Where is the chat in spanish please?

[trim osprey]

nvm foudn it https://docs.google.com/spreadsheets/d/1tAUaQrEHYgRsm1Lvrnj14HFHDwJWl0Bd9x0QePewNco/edit?gid=1227575351#gid=1227575351

[feral halo]

Ayo? @trim osprey level 1 !!!

[trim osprey]

LFGGG

[wide cargo]

Go to roles and look for the spanish role

-kaggle

[rare sorrel]

📘 Kaggle Notebooks

• Applio Notebook, by Vidal Kaggle
• Applio Notebook, by Shirou Kaggle
• Music Source Separation, by Shirou Kaggle
• UVR5 NO UI, by Eddy Kaggle
• RVC Mainline, by Hina Kaggle
• Original W-Okada's Voice Changer, Kaggle
• Modified W-Okada's Voice Changer, Kaggle
• 📖 How to use RVC Mainline Kaggle by Cauthess

Note: Kaggle limits GPU usage to 30 hours per week.

[wide cargo]

-okada

-colab

[rare sorrel]

☁️ Google Colabs

• Applio, by IA Hispano Google Colab
• RVC Disconnected, by Kit Lemonfoot Google Colab
• RVC Mainline, by Hina Google Colab
• AICoverGen-WebUI, by Hina Google Colab
• AICoverGen-NoWebUI [English], by Ardha, fixed by Eddy, Hina and Gdr Google Colab
• AICoverGen-NoWebUI [Spanish], by Eddy, Hina and Gdr Google Colab
• UVR5 NO UI, by Eddy Google Colab
• UVR5 UI, by Eddy Google Colab
• Modified W-Okada's Voice Changer, Google Colab

ℹ️ Note

While the Colab free plan provides up to 12 hours of daily usage, the GPU is typically available for only about 4 hours each day on average.

More commands: -audio, -hf, -local, -kaggle, -overtraining, -realtime, -rvc, -help

[polar flax]

I also used that a year ago, but it seems no longer updated and the models are old already, some are RVC v1 and more likely contains background noise and in worse quality

[viscid loom]

https://tenor.com/view/gojo-satoru-satoru-gojo-gojo-jujutsu-kaisen-manga-gif-17386688786442927012

[polar flax]

https://tenor.com/view/embed-fail-jujutsu-kaisen-gojo-gojo-satoru-gif-8338841269825890917

[ionic crystal]

2 steps ahead

[elder willow]

who can help me with a ddos

i have to fix this in rvc ai cover maker kaggle

[languid vault]

Hi everyone! I've just joined the server, nice to meet you guys!
I am an AI game creator, using the AI tool RPGGO. Try my game if you are interested!

sry it's not a self promo, I'll remove my URL here

[elder willow]

Ai will become a religious phenomenon. it is the one thing who is going to redeem us. no wonder i am working on this thing for full time.

(angle(ABC)=angle(BAC))
((angle(ACB)+angle(BAC)+angle(ABC))=180)
>>> equation angle(BAC)=angle(BCA)
(angle(ABC)=angle(BAC))
(angle(BAC)=angle(ACB))
((angle(ACB)+angle(BAC)+angle(ABC))=180)
>>> calculate
(angle(ACB)=angle(ABC))
(angle(BAC)=angle(ABC))
(angle(ACB)=60)
(angle(BAC)=angle(ACB))
(angle(ACB)=angle(BAC))
(angle(BAC)=60)
(angle(ABC)=angle(BAC))
(angle(ABC)=60)
((angle(ACB)+angle(BAC)+angle(ABC))=180)

end of program``` a demo of my Ai solving ncert class 9 questions.

the education system of any nation is a part of the religious phenomenons of it

even human language can be replicated by Ai

i have done a bit amount of work in computational english grammar

i just have to extend it more. i already extended, i was trying to understand what is lying in a communication between people.

machine learning could help. but i think rule based Ai has also to be tried. in some sense its more promising than machine learning in many aspects.

its 100% that rule based Ai is better than machine learning in Ais which solves systems.

not just human language replication, you need to replicate the systems humans created who value it so much.

i have built chemistry, math and geometry Ai.

[elder willow]

i just make one cover

lol

apparently with the nvidia p100 there are no errors.

but is slow

with an nvidia p100 gpu

no errors for now

[elder willow]

What can I do to make it use both gpu's? it seems to give an error

then I continue to fix the other failures with tta, uvr models, ect.

[earnest dragon]

Paid AI one sentence only needed to voice clone tts and lipsync ai the creator gave me pro for free. Here is a example I did.

[golden shard]

I know basic java but i want to know how can i include ,make or use AI model in java

[elfin hinge]

hey guys I have been researching tools for building datasets and found augmentoolkit. But I noticed the owner has basically nerfed/ filled the latest versions with bugs. My next step is I am going to try experimenting with some branches of his project in Gihub.

Would be interested to know if anyone has experience with this or other dataset tools. if not, who wants to join me on this task?

[feral halo]

Ayo? @elfin hinge level 1 !!!

[elfin hinge]

btw I spoke to an 'associate' of the founder of this tool and it seems they are making a commercial version and my guess is they nerfed the main open source one to 'encourage' non technical people to basically reach out to them for 'consulting'!

[polar flax]

local & online RVC applications that use gradio can provide API access

[elfin hinge]

it's also possble to just clone an older version of his tool in Git, as well as the branches idea. so I have a few avenues i am going down.

But I would be interested to be persuaded wheter it's a better idea to make my own dataset generating tool, and if so why?

[golden shard]

but they said make it in such a way that it is offline and doesnot require any net connectivity

[polar flax]

yeah the local one also provides API as well, through the localhost address

[wide cargo]

-kaggle

[rare sorrel]

📘 Kaggle Notebooks

• Applio Notebook, by Vidal Kaggle
• Applio Notebook, by Shirou Kaggle
• Music Source Separation, by Shirou Kaggle
• UVR5 NO UI, by Eddy Kaggle
• RVC Mainline, by Hina Kaggle
• Original W-Okada's Voice Changer, Kaggle
• Modified W-Okada's Voice Changer, Kaggle
• 📖 How to use RVC Mainline Kaggle by Cauthess

Note: Kaggle limits GPU usage to 30 hours per week.

[golden shard]

I dont get it , can you please explain more if possible

[polar flax]

if you dont know that thing, again better make a python program

[twilit furnace]

were can i have to know how to make a model?

[polar flax]

bonked him/her!

[earnest dragon]

English,Chinese,Japanese Lipysnc Version using only one sentence to clone the voice and it supports different languages.

[barren ocean]

dolboebii

[empty wave]

Hewwo chat

[untold tulip]

halooo

[patent quest]

hewu humans

i was trying to make a software and the AI constantly went down the wrong path, so went outside and engaged in human behaviour of smoking a cigarette

then I wrote a bit of code and asked AI to color the extra bits in

it did that quite well

I also bet the AI didn't do 20,000 steps yesterday and has sore calves

[feral halo]

Ayo? @patent quest level 5 !!!

[lone gust]

ı want voice model can ı help you

which model must use ?

[golden shard]

I will , can you suggest some programs to practice

[lone gust]

ı want use collab

[polar flax]

the python documentation itself

[covert lake]

u got a bad pc and want to make an rvc model on cloud right

[lone gust]

right

[covert lake]

whats ur pc gpu?

[viral bronze]

How i make dataset

[covert lake]

As you dont got a good PC, its better you use cloud for training an RVC Voice Model:

• Google Colabs (4 hours of daily gpu for free, not much hours, but easy to use):
• Applio (ui)
• Mainline (UI)
• RVCDISCONNECTED (no ui)
• Kaggles (a bit harder to use and needs phone number but gives 30 hours weekly of better gpus):
• Mainline (UI)
• Applio by Vidal (UI)
• Applio by Shirou (UI, no guide as of right now)

I'd personally suggest to use Kaggle

[twilit furnace]

Nvidia

[covert lake]

yea but which

[twilit furnace]

GTX 1660ti

[covert lake]

You can't train (make) a model on that locally

[twilit furnace]

Sad

[covert lake]

You can only inference (use models) locally, but for training you can just do it on cloud

As you dont got a good PC, its better you use cloud for training an RVC Voice Model:

• Google Colabs (4 hours of daily gpu for free, not much hours, but easy to use):
• Applio (ui)
• Mainline (UI)
• RVCDISCONNECTED (no ui)
• Kaggles (a bit harder to use and needs phone number but gives 30 hours weekly of better gpus):
• Mainline (UI)
• Applio by Vidal (UI)
• Applio by Shirou (UI, no guide as of right now)

I'd suggest kaggle

[twilit furnace]

I will try it thanks u

[feral halo]

Ayo? @twilit furnace level 2 !!!

[covert lake]

yw

[covert lake]

are you making your own version of applio?

[covert lake]

https://docs.aihub.wtf/rvc/resources/datasets/

[cedar juniper]

hi !

can i use rvc with a m1 mac ?

i got a link but there is only amd & nvidia card

[fleet crag]

Good morning

I have just joined this discord server because I am new to AI and I want to inform myself a little, I have downloaded the LM Studio program and it is awesome, I have been playing a little with a 10GB model that I downloaded called llama 3. 1 7b 8 QBIT but there are many things that escape me, I know that the model is loaded in ram memory and that is important for it to work fast but there are things I do not understand as the tokens and context, I have noticed that the default context was 4096 but speaking the model has increased to 8100 but I can also set that value in the settings up to 132,000 which is much, someone can explain all this to me? thanks

[covert lake]

technically but only for inference (use models) and would have to follow from the github of the original rvc https://github.com/RVC-Project/Retrieval-based-Voice-Conversion-WebUI/blob/main/docs/en/README.en.md#preparing-the-environment

Personally i'd suggest just doing it on cloud where u can inference and train (make models) faster

[cedar juniper]

that u for the answer u goated don't forget that

[golden shard]

oh i know python at intermediate level

[covert lake]

your welcome

[ionic pumice]

https://x.com/SSeeley92/status/1832909031096099267

🥀 Sarah 🥀 (@SSeeley92) on X

Good wholesome fun for the kids. 😁🤘🏼🖤

Twitter · Sep 8, 2024, 10:29 PM

slay

[dreamy heath]

hi

[serene panther]

hey is anyone able to send me an AI model of Cream heart who is Button Mash's Mom?

[feral halo]

Ayo? @serene panther level 4 !!!

[boreal trail]

л

[serene panther]

uhm i speak english

[covert lake]

You can search rvc ai voice model at:

• #1175430844685484042
• #🔍│find-models
• https://weights.gg/ (login required)
• https://huggingface.co/models (but watch out cus in hugging face there arent only rvc ai voice models)
• https://applio.org/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.aihub.wtf/essentials/how-to-make-voice-models/

[fallen ruin]

https://tenor.com/view/rengoku-good-morning-gif-10721789691834654422

[dreamy heath]

https://tenor.com/view/donuts-sprinkles-delicious-gif-17462016

[fallen ruin]

https://tenor.com/view/waking-up-patrick-oh-boy-1pm-gif-26372839

[fallen ruin]

https://tenor.com/view/donuts-sprinkles-delicious-gif-17462016

[dreamy heath]

https://tenor.com/view/rengoku-rengoku-donut-demon-kny-kimetsu-no-yaiba-gif-25040774

[fallen ruin]

😭

[fallen ruin]

https://tenor.com/view/rengoku-rengoku-donut-demon-kny-kimetsu-no-yaiba-gif-25040774

[feral halo]

Ayo? @dreamy heath level 2 !!!

[fallen ruin]

https://tenor.com/view/one-piece-fire-ace-anime-portgas-d-ace-gif-17433525

https://tenor.com/view/fosca-flamejante-donut-man-gif-11989025

https://tenor.com/view/demon-slayer-giyuu-gif-23140928

Baker akaza 🍩🥯

[dreamy heath]

🍩

[fallen ruin]

🥯

https://tenor.com/view/bagel-excited-jensen-ackles-dancing-dance-gif-13024208

[dreamy heath]

🏀

[dreamy heath]

🍩

[fallen ruin]

😼

🥯

https://tenor.com/view/sonic-bagelnet-bagel-bagels-gif-15161223

[iron aspen]

Hi, how we do ai voice with our voice?

[worldly phoenix]

Hi everyone, nice to be here to learn from you all.

[serene panther]

@covert lake man i'm guessing it was removed or something cause voice actor doesn't want her voice being ai cloned

[polar flax]

https://tenor.com/view/furina-embed-fail-gif-10775077421402251353

[fallen ruin]

https://tenor.com/view/muzan-demon-slayer-muzan-kibutsuji-muzan-demon-slayer-ds-gif-10446059776737452225

https://tenor.com/view/michael-jackson-popcorn-gif-6840783197839228593

[polar flax]

read the guide to get started https://docs.aihub.wtf/

[tawny owl]

Hi guys
Is it possible to upload my Voice and Instrumental to Suno AI to make music?

[fallen ruin]

https://tenor.com/view/jojo-jojo-reference-oh-shit-boi-is-this-a-motherfucking-jojos-reference-gif-17147476

[covert lake]

i dunno abt that creator but ig thats the reason then sans

promo in #1159290752195633273

[stoic ibex]

its my fifth time joining here and no one ever replied

its my birthday, could u help me creating a model please

[gilded frigate]

Hello, who has a sound identifier ai? Like if I have a sound of something that the ai detects what it is

[novel nymph]

🔥😎

[novel nymph]

No

[novel nymph]

No one has anything like that

[gilded frigate]

How? Why does something like that not exist? It’s just like Shazam but than from other sounds not music

[stoic ibex]

ableton has that

but for samples 🤣

[gilded frigate]

What kind of samples

[feral halo]

Ayo? @gilded frigate level 1 !!!

[gilded frigate]

@stoic ibex

[wanton mirage]

Hi! Does anyone know how a singer can sing the same song but in Spanish?

because I have seen several covers of English music sung in Spanish

[cunning dagger]

türkçeye uygun kadın ses modeli olan var mı aramızda arkadaşlar

[novel nymph]

#1159595680659157132

[obtuse cave]

merhaba

kemal sunal lazım bana var mıdır

[steady glacier]

beta varmı aramızda

[dapper island]

Are there any living people here?

[polar flax]

https://tenor.com/view/dead-my-honest-reaction-damn-skull-skeleton-gif-15885937969029652188

[solar zealot]

Are there any means to create voice models for free without logging in?

[dapper island]

is it possible to write here what is allowed on other servers?

There's just something...

[ionic pumice]

you can try local if u have a good enough computer specifications

[solar zealot]

Local what?

[feral halo]

Ayo? @solar zealot level 1 !!!

[dapper island]

well, the last letter of the Latin alphabet and "we don't leave our own" in Russian.

[ionic pumice]

local training

[solar zealot]

Aaaand....how do I do that?

Is there a software for it?

[ionic pumice]

-local

[rare sorrel]

🖥️ Local stuff

• 🍏 Applio, by IA Hispano GitHub
• Mangio-RVC-Fork, by Mangio621 Huggingface
• RVC Studio, by SayanoAI Huggingface
• AICoverGen, by SociallyIneptWeeb GitHub
• Replay, by Replay Team Website
• Original RVC, by the RVC-Project team GitHub
• GPT-SoVITS, by RVC-Boss GitHub

Credits to Faze Masta and Antasma for compiling these links.

More commands: -audio, -colabs, -overtrain, -realtime, -rvc, -help

[solar zealot]

Which one of these is for voice cloning?

-help

[rare sorrel]

Command has expired

Overtraining

You can detect if a model is overtraining if the TensorBoard graph starts to rise and never comes back down. An overtrained model will sound robotic, muffled, and won't be able to articulate words well.

Check these resources to learn more about this topic

• Epochs & TensorBoard from AI HUB Docs
• TensorBoard from 🍏 Applio Docs

[ionic pumice]

check out the applio or original rvc link

[timid musk]

mongus

[fallen ruin]

Literally MK 😂

[dreamy heath]

hi

[mystic frost]

Hi

[dreamy heath]

literally me

[dreamy heath]

how are you

[mystic frost]

Good

[feral halo]

Ayo? @mystic frost level 1 !!!

[mystic frost]

🤑

Yeah

[floral bay]

yall how do i get the output of this covergen no ui thingy

its so confusing

[elder willow]

port of rvc ai cover maker

[somber kindle]

To be fair, you have to have a very high IQ to understand Rick and Morty. The humor is extremely subtle, and without a solid grasp of theoretical physics most of the jokes will go over a typical viewer's head. There's also Rick's nihilistic outlook, which is deftly woven into his characterisation - his personal philosophy draws heavily from Narodnaya Volya literature, for instance. The fans understand this stuff; they have the intellectual capacity to truly appreciate the depths of these jokes, to realize that they're not just funny- they say something deep about LIFE. As a consequence people who dislike Rick and Morty truly ARE idiots- of course they wouldn't appreciate, for instance, the humour in Rick's existencial catchphrase "Wubba Lubba Dub Dub," which itself is a cryptic reference to Turgenev's Russian epic Fathers and Sons. I'm smirking right now just imagining one of those addlepated simpletons scratching their heads in confusion as Dan Harmon's genius unfolds itself on their television screens. What fools... how I pity them. 😂 And yes by the way, I DO have a Rick and Morty tattoo. And no, you cannot see it. It's for the ladies' eyes only- And even they have to demonstrate that they're within 5 IQ points of my own (preferably lower) beforehand.

[somber kindle]

i searched this up on google and it gave me a wikipedia page that said "intellectual giftedness"

im not sure how to feel about that

[fallen ruin]

https://tenor.com/view/explosion-big-explosion-kimetsu-no-yaiba-ubuyashiki-kny-gif-15737026962177320088

[novel nymph]

😳🤯

[humble minnow]

a real human sung it and someone just converted it with AI

[fallen ruin]

Muzan Jackson during the explosion

[dreamy heath]

boom

[dreamy heath]

[dreamy heath]

real

[covert lake]

wait how does it not work with t4x2 😭

[polar flax]

https://tenor.com/view/exploding-car-explode-cat-boom-bah-gif-5387928273689188008

[elder willow]

I'm trying to fix that

[covert lake]

idk if it can help but for my own flux kaggle notebook to be sure it used both gpus i did:

if torch.cuda.is_available():
    device = "cuda"
    print("Using GPU:", device)
    # Get the number of available Cuda GPUs
    n_gpu = torch.cuda.device_count()
    print(f"Number of GPUs available: {n_gpu}")
else:
    device = "cpu"
    print("Using CPU:", device)

[covert lake]

btw whats ur pc gpu

[elder willow]

you gave me an idea

for my port of uvr

I would say my uvr5 ui port is ready

[dense ruin]

wsup

[covert lake]

mm what is it?

[elder willow]

integrate that into uvr

does not use the two t4

[gentle canopy]

hi

[covert lake]

alright

the full version has also pytorch xla for tpus btw

but its slower than gpu on pytorch

[elder willow]

I'm fighting this bug in my rvc ai cover maker port

[charred python]

Hi

[elder willow]

happens after making a few inferences with a p100 gpu

[covert lake]

seems caused by https://github.com/ShiromiyaG/RVC-AI-Cover-Maker-UI/blob/7ce3eeb17342a06ef7071cf399c070565639bb15/programs/music_separation_code/utils.py#L99

[elder willow]

and yesterday this happened to me:

It happens when you want to use both T4 GPUs.

[naive sonnet]

cake disgusted ai

[upper fable]

Why on the illyria zero hugging face I can’t put the bpm I want on my voice?

[elder willow]

😋

[fallen ruin]

https://discord.com/channels/1159260121998827560/1159290193619189821

[tiny sandal]

Anyone can recomend me a good voice modifier?

I can't found a good one

[upper fable]

Any good Daniel Caesar models?

[dreamy heath]

[elder willow]

I want to eat that cake, the cockroaches add a special touch to it 🥰 😋

hi

[fallen ruin]

All bugs fear this monster

[earnest dragon]

for tts or what

[random mason]

Hi,
Got a bit of a odd request.
I'm looking for some hilariously terrible free voice generators for shitposting.
Would anyone be able to point me in the right direction?
Cheers

[upper fable]

Nah for making a song

[gilded crypt]

can anyone tell me where i can go to make and train my own voice for voice changer Client i wanna make my own one

[earnest dragon]

then the bpm of voice will be the input audios speed

[upper fable]

How do I make it adjust it because it sounds off

I remember doing it back in collab and it worked soo well

[dark harbor]

tring to find some voice moddels

[solar zealot]

Integrated graphics

[sturdy elbow]

Which model does this happen with?

Btw, it only uses two GPUs when a BS Roformer-based model is used

[elder willow]

default

and denoise

[lapis prawn]

i want to train a voice model and have an amd card. What can i use?

[elder willow]

and bs roformer

[sturdy elbow]

Okay, I'll fix it

[elder willow]

and fix this

[upper fable]

I need the bpm to work

[elder willow]

Apparently if you move the audio settings a little it breaks

[lapis prawn]

i want to train a voice model and have an amd card. What can i use?

[ionic crystal]

Wtf dead server

[sturdy elbow]

I believe it's been fixed on the T4s, I'll check on the p100s

[elder willow]

It occurs to me to put this: use_amp = getattr(config.training, 'use_amp', False) # default to False if not present

[covert lake]

hell no don't follow that advice then u cant use it locally

u have to do it on cloud

[sturdy elbow]

Hmmm that solved the problem on the p100?

[dreamy heath]

hi

[earnest dragon]

if it sounds off then the vocals are not the same length as the insturmentals

[elder willow]

i will check

[covert lake]

As you dont got a good PC, its better you use cloud for training an RVC Voice Model:

• Google Colabs (4 hours of daily gpu for free, not much hours, but easy to use):
• Applio (ui)
• Mainline (UI)
• RVCDISCONNECTED (no ui)
• Kaggles (a bit harder to use and needs phone number but gives 30 hours weekly of better gpus):
• Mainline (UI)
• Applio by Vidal (UI)
• Applio by Shirou (UI, no guide as of right now)

But there is no way without loggin in

[ionic crystal]

@covert lake

[covert lake]

on windows:
You can use amd locally but you need zluda. Here is a applio pr talking about it https://github.com/IAHispano/Applio/pull/513,
You can find https://github.com/IAHispano/Applio/blob/main/assets/zluda/README.md how to install it

[elder willow]

i have an guide for applio kaggle by vidal: https://docs.google.com/document/d/1LXaZpb446wK9gTeMiTKwbB8xQVrPMPovK7aNiE6oh-0/edit#heading=h.zhj4p2wbtlbn

[lapis prawn]

i will try thank you

[earnest dragon]

Zluda does not work on ai like this for training.

[covert lake]

wasn't that the same guide as https://applio.org/guides/58 ? also the one without guide is Shirou one

[elder willow]

I did that to make deivih happy

The one I have on Google Docs is more up to date

[lapis prawn]

so i cant train my own voice model with an amd card?

[feral halo]

Ayo? @lapis prawn level 1 !!!

[lapis prawn]

i have an old 1080 i could use

[ionic crystal]

Dont

[covert lake]

I don't use zluda as im on integrated graphics and i do things on cloud,
But based on what @night lake said yea #✨│ai-help message

[elder willow]

use this

https://github.com/IAHispano/Applio/tree/exp/cpu-training

[earnest dragon]

No you can't unless Linux

[earnest dragon]

So slow

[covert lake]

cpu training is slow asf though

[elder willow]

but works

[earnest dragon]

He should just use his gtx 1080

[lapis prawn]

what would i use on linux?

[covert lake]

its just applio that added that fake gpu thing for some reasons

But the commits about that cpu training doesn't make cpu training faster

[covert lake]

u can with zluda technically

[earnest dragon]

the same stuff but the Linux version and rocm and stuff on Linux there is probably a guide for it

[covert lake]

its basically cuda for amd

[earnest dragon]

No does applio use cudnn

[ionic crystal]

@covert lake get a job

[lapis prawn]

fuck amd

xD

[ionic crystal]

you confuse zluda with rocm

Ive never even heard the word zluda before

[covert lake]

what?? no

[feral halo]

Ayo? @covert lake level 103 !!!

[covert lake]

https://github.com/vosen/ZLUDA

[ionic crystal]

Cuda on ???

[covert lake]

applio has working support with zluda seen in https://github.com/IAHispano/Applio/pull/593 and https://github.com/IAHispano/Applio/pull/513

[covert lake]

yes cuda on amd gpus

[ionic crystal]

Thats for intel you idiot

[covert lake]

🤦‍♂️ its for amd too

[elder willow]

I still remember the day that AMD banned Zluda from their GPUs

[ionic crystal]

Oh it is?

This is my most honest reaction i could ever give

[lapis prawn]

im cooked

[ionic crystal]

Yes

[elder willow]

lol

[ionic crystal]

9 ago

[covert lake]

ye lol

[covert lake]

https://github.com/vosen/ZLUDA/issues/234

[ionic crystal]

@covert lake

[elder willow]

OMG, my port works

[ionic crystal]

What does arm and qualcomm have to do with amd

[ionic crystal]

Okay and

It looks kinda ugly.

https://tenor.com/view/fnaf-fnaf-security-breach-purple-guy-man-behind-the-slaughter-five-nights-at-freddys-gif-24196550

[elder willow]

mmm

[covert lake]

ay good

[ionic crystal]

Fredy hur hur hur

[covert lake]

zluda works for amd

[elder willow]

is the default theme of rvc ai cover maker

[covert lake]

why even spread negativity lol

[dreamy heath]

hi

[feral halo]

Ayo? @dreamy heath level 3 !!!

[covert lake]

i feel like its just a troll atp

[dreamy heath]

[ionic crystal]

Bro its purple

Thats not great in eyes

[elder willow]

https://tenor.com/view/purple-guy-dance-william-afton-dance-purple-guy-jiggy-purple-jiggy-purple-dance-gif-17627065305568870467

[ionic crystal]

Brown would be better ngl

Or gray

[covert lake]

https://tenor.com/view/opinion-gif-20305030

[sturdy elbow]

Yeah, I couldn't find a better theme lol But I'll add another one

[sturdy elbow]

lol

[ionic crystal]

DU KLEINES GRÜNES MÄNNCHEN

Schaß manschgal he

[sick vapor]

https://tenor.com/view/butt-butthole-tight-tickle-legs-gif-15619193

[covert lake]

whatever ur smoking is good

[ionic crystal]

Translated wrong lmfao

It means green little figure

[proud bloom]

hi

[ionic crystal]

@proud bloomno

Ur pfp says it all about you

Let me gueds ur trying to finda model of N

[covert lake]

i personally like it

[ionic crystal]

https://pokemon.fandom.com/wiki/N

[elder willow]

"NoCrypt/miku@1.2.2"

[dreamy heath]

hi

[sturdy elbow]

Thank you, that relieves me lol

[dreamy heath]

hello

[covert lake]

wtf i was looking at that too bc all other themes in https://huggingface.co/spaces/gradio/theme-gallery are in error

[elder willow]

are broken

[covert lake]

ye fr idk why

[sturdy elbow]

Yes, that's why I didn't choose another one...

[covert lake]

huggingface is on drugs

[sturdy elbow]

Hmmmm no?

[ionic crystal]

@covert lake i wonder what happens if nvidia suddenly decides to destroy all their drivers as a experiment

[feral halo]

Ayo? @ionic crystal level 27 !!!

[covert lake]

https://tenor.com/view/two-steps-ahead-im-always-two-steps-ahead-nikado-avocado-nikado-avacado-gif-24881891

[earnest dragon]

No Intel they got rid of and only works on amd

Hyphonix Lipsync RVC Cover https://autocompressor.net/av1?s=KoTVeGfm

[ionic crystal]

Nvidia decides to make their gpus catch on fire with a new driver that makes all gpus go to 10000w

And overclocks by 5v

[covert lake]

https://tenor.com/view/madagscar-penguins-kaboom-gif-9833865

[ionic crystal]

@covert lake intel decides to shut their servers down

[earnest dragon]

It's funny a 3080ti can pull more power then a 3090 if you use the official 1000w bios it has

[ionic crystal]

Reta-

Rita

[elder willow]

we need create an Ai Hub chan gradio theme

[ionic crystal]

But ai hub is dead

[elder willow]

In that case, this server is for empanadas and cheese

[ionic crystal]

[sturdy elbow]

@elder willow Did that change make it work then?

[ionic crystal]

@elder willow suck on ball

Literally

https://www.ebay.com/itm/325795828709?norover=1&mkevt=1&mkrid=711-169407-643243-7&mkcid=2&itemid=325795828709&targetid=293946777986&device=m&mktype=pla&googleloc=9062728&poi=&campaignid=20529249311&mkgroupid=157138816230&rlsatarget=pla-293946777986&abcId=&merchantid=5399977204&gad_source=1&gclid=CjwKCAjw3P-2BhAEEiwA3yPhwCovWvz7gaPcsru7EXLpZHskPhyiQ01CJdzy4gZuT6IQKZqR3YLnlRoCf0AQAvD_BwE

[elder willow]

i have a fork for rvc ai cover maker

[ionic crystal]

https://m.made-in-china.com/product/Bright-Surface-Wolfram-Ball-with-Diameter-5mm-for-Hunting-Usage-764683472.html

[elder willow]

dont enter

[ionic crystal]

I did already

[ionic crystal]

Dont enter

https://m.made-in-china.com/product/Bright-Surface-Wolfram-Ball-with-Diameter-5mm-for-Hunting-Usage-764683472.html

[sturdy elbow]

Aaahh i see, I made a change to the code, now it works on the p100, i think

[ionic crystal]

@sturdy elbow

Fuck i need this in german

[earnest dragon]

https://www.techpowerup.com/vgabios/235179/235179 1000w 3080ti bios

[sturdy elbow]

?

[ionic crystal]

Ok

[earnest dragon]

Yeah and the power limit slider is accurate so you can set it to a percent and it will do that percentage of 1000w so you can have it and not go full 1000w

[ionic crystal]

For?

[earnest dragon]

Switch is not region locked

[tiny sandal]

Hey can somebody recomend me a good voice modifier pls

[ionic crystal]

They discontinued the game series in all countries except japan

The last game was 3 and for 3ds

[solar zealot]

Applio doesn't look like you can upload sample audio to train the voice model

[ionic crystal]

They even took 3 years to get the game into germany

2nd game was released 2014 in japan and 2017 in germany

[tiny sandal]

Hey can somebody recomend me a good voice modifier plss

[feral halo]

Ayo? @tiny sandal level 1 !!!

[ionic crystal]

No

[earnest dragon]

Okay then get Japan copy

[tiny sandal]

:c

Thx

[ionic crystal]

Yeah im gonna learn japanese totally

[earnest dragon]

subtitles if game has them

[pale patrol]

well good luck its gonna take u months or a year

[ionic crystal]

Not doing that

[pale patrol]

lol

[ionic crystal]

Since when do japanese copies have subtitles

[earnest dragon]

some games have subtitles when people are talking for accessibility

[wide cargo]

Which one

[earnest dragon]

Brawl Stars overrated

[dreamy heath]

piper😻

[wide cargo]

Shush

[pale patrol]

its true

[earnest dragon]

It's the truth only kids in school play it.

[pale patrol]

good thing i dont play it

[ionic crystal]

Only????

It took me like 9 years to even learn english

[pale patrol]

it took me 4 years to speak english :p

[earnest dragon]

Yeah when I was in school I use to play black ops 2 and 3 using this one gaming remote desktop app that let me connect to my pc from the school chromebook browser

[solar zealot]

Apparently drive-based voice training isn't my forte

[pale patrol]

was it laggy tho?

[earnest dragon]

What are you looking for then

[earnest dragon]

No it wasn't because my home internet is 1000 1000

[feral halo]

Ayo? @solar zealot level 2 !!!

[solar zealot]

Way to train a voice model from an audio file. Preferably a way to do it for free, without logging in and without Collab or other drive-based thingamajigbomb

Or a software where that is possible

[earnest dragon]

what gpu do you have

[pale patrol]

dangggg.

[solar zealot]

I already answered that

Integrated graphics

[earnest dragon]

where

[pale patrol]

o-o

[earnest dragon]

Then those are the only options you gave.