🧬│ai-chat

[versed jetty]

iam cookie monster not you🤣

[storm meadow]

https://docs.applio.org

[sudden temple]

I use AIVA, need license

[valid scroll]

#🔍│help-w-okada message

[storm meadow]

Some stuff is a bit outdated or inaccurate tho on applios documentation tho it should work like RVC since it’s a fork of it

[lean blaze]

.wk

[indigo pike]

xaviersobased in AI HUB

👑  ⁦Adiiose⁩ - 4206 plays
 2.  ⁦remy⁩ - 3988 plays
 3.  ⁦exan⁩ - 3141 plays
 4.  ⁦luminal⁩ - 2823 plays
 5.  ⁦yeresin⁩ - 2299 plays
 6.  ⁦uh1c⁩ - 990 plays
 7.  ⁦sqwab⁩ - 947 plays
 8.  ⁦mox⁩ - 862 plays
 9.  ⁦gtemq⁩ - 767 plays
 10. ⁦osc⁩ - 765 plays
 11.  ⁦buga⁩ - 671 plays
** 12.**  ⁦gilbert⁩ - 658 plays
 13.  ⁦cxvz⁩ - 615 plays
 14.  ⁦Frankie⁩ - 603 plays

cloud rap - dark plugg - sigilkore Artist - 296 listeners - 40451 plays - 136 avg

[storm meadow]

Ai voice actor?

[versed jetty]

dexter

[leaden basin]

lol sorry

[leaden basin]

okok

[versed jetty]

iam just kidding

[lean blaze]

.wkt

[indigo pike]

ghost town (444jet n exset) by xaviersobased in AI HUB

1.  ⁦gilbert⁩ - 64 plays

[sudden temple]

@storm meadow create sample and music

And you ?

Are you french ?

[storm meadow]

No https://discord.com/channels/1159260121998827560/1159520446799155350

[sudden temple]

Okay

[feral halo]

Ayo? @sudden temple level 1 !!!

[versed jetty]

shut up app you are f@##$

[steel coral]

Bro really censorin himself

[versed jetty]

yes he is a fu#$%in app

[storm meadow]

Bro you are allowed to swear here

[steel coral]

boy shut the hell up 🤫

goofy ahh

lmao

[versed jetty]

iam just kidding ok i will stop this iam very polite not bad dont worry

[wintry sinew]

dont fucking swear 😡

[steel coral]

mustofa eto faizlami korio na (mostofa dont play so much)

[versed jetty]

who is mostofa

[steel coral]

mustofa

[versed jetty]

ok

[steel coral]

moustafa

never saw a person spell it that way

[versed jetty]

iam not moustafa

[versed jetty]

no man

[steel coral]

[feral halo]

Ayo? @steel coral level 2 !!!

[steel coral]

anhoinig ahh boyt

[sudden timber]

quick question, how do i use files from weights.gg in python? not a web ui

[versed jetty]

copy it

[rocky ice]

:)

[feral halo]

Ayo? @rocky ice level 2 !!!

[versed jetty]

🤣 ]

[feral halo]

Ayo? @versed jetty level 4 !!!

[elder willow]

made a model request TriHard 7

[versed jetty]

yes

[sudden timber]

i never knew how to do that, thanks for the great advice

i need actual help tho, any python library that works

[storm meadow]

#✨│ai-help #🔊│ai-development

[rocky ice]

Hello

[opal ginkgo]

Have you seen 🐢 ?

[whole drum]

yo

anyone know where to get RatIRL voicemodel from?

[storm meadow]

@minor blade have you?

[storm meadow]

#1175430844685484042

[minor blade]

🐢

[whole drum]

not there

[opal ginkgo]

There he is.

Thanks.

[storm meadow]

weights.gg

[versed jetty]

yes i have one

[whole drum]

ty

[feral halo]

Ayo? @whole drum level 1 !!!

[versed jetty]

Ayo?@arclame level 0 !!!

[acoustic maple]

what's a believable female voice model?

[muted fable]

Running on local URL: http://127.0.0.1:6969

Could not create share link. Missing file: /usr/local/lib/python3.10/dist-packages/gradio/frpc_linux_amd64_v0.2.

Please check your internet connection. This can happen if your antivirus software blocks the download of this file. You can install manually by following these steps:

1. Download this file: https://cdn-media.huggingface.co/frpc-gradio-0.2/frpc_linux_amd64
2. Rename the downloaded file to: frpc_linux_amd64_v0.2
3. Move the file to this location: /usr/local/lib/python3.10/dist-packages/gradio

not again , they say it was fixed😭

[ebon scaffold]

is there an ai that can change the lyrics of a song?

[covert lake]

retry now?

[muted fable]

okay bro im trying

[muted fable]

bro i tried twice and the result is the same. i even tried it with different accounts but didn't work. what shall we do ?

[covert lake]

seems like gradio still having some problems

[muted fable]

running on colab is very easy , fast and powerful😭

[muted fable]

will it be fixed soon?

[covert lake]

gradio is having problems the whole day
so idk

[muted fable]

but can we follow the rules

Please check your internet connection. This can happen if your antivirus software blocks the download of this file. You can install manually by following these steps:

Download this file: https://cdn-media.huggingface.co/frpc-gradio-0.2/frpc_linux_amd64
Rename the downloaded file to: frpc_linux_amd64_v0.2
Move the file to this location: /usr/local/lib/python3.10/dist-packages/gradio

this one? can we do it?

[covert lake]

nope as its the online public link, you can only just wait for the api to be back up

[hidden cargo]

could anyone here help me with an opinion on a website? If it looks sketchy and if so why exactly. I found a website and want to order from it but it's quite new and I want to hear other people's opinions on it and I didn't know who to ask so I randomly press on this server

[muted fable]

ill wait bro

[forest quarry]

yeah

[fallen ruin]

YeAh BeCaUsE YoU’rE UnFuNnY

YeS, It'S BeEn 23 YeArS SiNcE ThE AtTaCkS, AnD YeS, SoMe PeOpLe CaN MaKe JoKeS AbOuT It.

BuT ThAt Is JuSt In PoOr TaStE, ThAt WaSn'T EvEn A JokE. ThAt WaS YoU SeEiNg A PiCtUrE Of ThE ToWeRs, DiVoRcEd FrOm ThE EvEnT, AnD YoU GoInG OuT Of YoUr WaY To SaY "hey guys, 9/11, funny right?"

WhAt Is ThErE To ExPlAiN. YoU MaDe A fuCkInG 9/11 JoKe On A PoSt ThAt HaD NoThInG To Do WiTh 9/11.

My GuY. YoU KeEp GeTtInG ToLd ThE ReAsOn. YoU MaDe An UnCaLlEd FoR AnD DeEpLy UnFuNnY 9/11 JoKe JuSt BeCaUsE YoU SaW An ImAgE Of ThE WoRlD TrAdE CeNtEr.

ThE OnLy ReAsOn I'm StIlL GiViNg ThIs ThE TiMe Of DaY Is BeCaUsE ThErE's A ChAnCe YoU'rE JuSt A KiD WhO GeNuInElY DoEsN't SeE ThE PrObLeM.

BuT MoRe ThAn LiKeLy, ThIs Is JuSt RaGeBaIt, BeCaUsE HoLy ShIt YoU CaNnOt Be ThIs ToNe DeAf.

[light bone]

hi hub ai

great cumback

[elder willow]

🥰

[gentle canopy]

hi

[light bone]

can i get trusted member role / thing to talk in #✦│chat and model maker

[wintry sinew]

done

[feral halo]

Vote Booster: Vote now for a 10% boost. https://arcane.bot/vote

Rewards for **AI HUB**:

Level 1: <@&1159345504442925056>
Level 5: <@&1159520205395984475>
Level 10: <@&1159537820352917516>
Level 20: <@&1159537862115598396>
Level 30: <@&1159537896471138345>
Level 40: <@&1159537927278309477>
Level 50: <@&1159537986053099542>
Level 60: <@&1159538018206625932>
Level 70: <@&1159538074439663666>
Level 80: <@&1159538136087527504>
Level 90: <@&1159538241897242664>
Level 100: <@&1159538281512439920>
Level 110: <@&1159538318183243777>
Level 120: <@&1159538343986614317>
Level 130: <@&1159538478669901844>

[light bone]

someone is crazy

[wintry sinew]

IM NOT

CRAZY

[wintry sinew]

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

[elder willow]

-guide

[rare sorrel]

Documentation

📚 All-In-One English documentation

https://docs.aihub.wtf/

❔ English FAQ

https://docs.aihub.wtf/faq

Credits: AI HUB FRANCE

✨ More guides

Full AI Voice Model Training Guide (Local)

Link: YouTube
Credits: Christopher Villanueva

Model training with Mainline RVC

Link: Rentry
credits: Raven (ravencutie21)

AICoverGen Colab Guide

Link: Google Docs
Credits: Eddy (Spanish Helper)

Create a model with RVC disconnected (colab)

Link: Google Docs
Credits: Angetyde

How To Make an AI Cover With Ilaria RVC

Link: Rentry
Credits: 👽 Julia (ailen2091)

[versed jetty]

hi

how can i find online web for learning python language in games

........................................

iam monster

,,

[rapid oracle]

Hi everyone!
Happy to be part of this community. I'm an electronics engineer that learned how to code in Python with MIT MooC some years ago. I recently took a course in Applied Data Science that gave lots of perspective so I'd like to start building my own applications especially in SaaS space. I coded a couple of websites in HTML, CSS, PHP, MySQL so the basic package but not too much into JS. I used WF and Figma for my last website. So any recommendation either application, iDE, or prog language to get closer to build something on my own will be welcome

[covert lake]

ur back?

[light bone]

yuhh

[versed jetty]

@rapid oracle please
how can i find online web for learning python language in games

[light bone]

i found a lot of studio sessions i must make new models

[versed jetty]

how can i find online web for learning python language in games

how can i find online web for learning python language in games

how can i find online web for learning python language in games

how can i find online web for learning python language in games

[light bone]

i think https://www.google.pl/?hl=pl

[versed jetty]

you talking for me

[elder willow]

https://letmegooglethat.com/?q=pygame+on+the+web

[light bone]

no to myself

[rapid oracle]

not sure. There are lots in coursera, plurasight or YT but probably someone here has a more curated view on the most effective to get started. I don't

[versed jetty]

give me name of any web please man

[light bone]

just use google

[elder willow]

google it bro

https://letmegooglethat.com/?q=pygame+on+the+web

[versed jetty]

ok iam sorry but i think google answers is stupid any way i will try google

[elder willow]

google literally has ai in the search page now

google's actuallly really smart

[heavy summit]

How do i make an ai cover? Help...TvT

[tacit hearth]

@elder willow dm bakar mısın?

[covert lake]

[feral halo]

Ayo? @covert lake level 86 !!!

[covert lake]

https://docs.aihub.wtf

[fringe badge]

anyone know why whenever i open start_http it closes

[elder willow]

guys, do you guys know how to fix Okada's command prom pop out before disappearing in seconds?

[feral halo]

Ayo? @elder willow level 1 !!!

[elder willow]

https://tenor.com/view/we-serve-food-here-sir-squidward-tentacles-bubble-bass-spongebob-squarepants-order-something-normal-gif-17867162

We do AI content here sir

Even then if we were to submit, the submission wouldn’t be under your name

[fringe badge]

i need help with that too

[fringe badge]

i got it working

[feral halo]

Ayo? @fringe badge level 1 !!!

[elder willow]

how?!

[fringe badge]

i downloaded the wrong file since i have a AMD cpu

[elder willow]

ohjhh

how do i check my cpu?!

[fringe badge]

open task manager

and go to performances

and it should show ur gpu and cpu

[exotic osprey]

.

[elder willow]

mine is 13th gen intel(R) Core(TM) i7-13700H

[elder willow]

so TM right?

[fringe badge]

so u have a intel

[fringe badge]

If you’re on Windows and have an AMD / Intel ARC GPU, you need the win_onnxdirectML-cuda version of the software

dm me ill send u it

[elder willow]

i have geforce rtx and Inte;l sooo Intel ARC GPU right?

[fringe badge]

ya

[elder willow]

oki!

i'll try it out

[remote thunder]

hi

how do i access all sound models

[elder willow]

can you send me the file hehe-

i'm confused on finding it

[fringe badge]

sure dm

[elder willow]

i can't dm you for some reasson

we have to be friend first?

[honest mantle]

I have an issue ,on google colabs ,this link ,it does not work the public url,why?: https://colab.research.google.com/github/Blane187/RVC-notebook/blob/main/Easy_GUI.ipynb

[quartz roost]

-realtime

[rare sorrel]

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

[honest mantle]

I used to do it from phone

[amber light]

probably old

-colab

[rare sorrel]

☁️ Google Colabs

• Applio, by IA Hispano Google Colab
• AICoverGen-WebUI, modded by Hina Google Colab
• AICoverGen-NoWebUI [English], by Ardha, fixed by Eddy, Hina and Gdr Google Colab
• AICoverGen-NoWebUI [Spanish], credits to Eddy, Hina and Gdr for translating and fixing Google Colab
• Ilaria RVC, by thestingerx Google Colab
• RVC Disconnected, by Kit Lemonfoot Google Colab
• easyGUI, by rejects Google Colab
• 🆕 UVR5 NO UI for Google Colab, by Eddy Google Colab

How to use?

🤗 Hugginface Spaces

• Applio, by IA Hispano Huggingface Spaces
• Ilaria RVC, by thestingerx Hugginface Spaces
• RVC-HFv2, by r3gm Huggingface Spaces
• AICoverGen, by r3gm Huggingface Spaces
• Advanced RVC Inference, by r3gm Huggingface Spaces
• RVC v2 Huggingface version, by Clebersla Huggingface Spaces

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

[honest mantle]

Look : https://colab.research.google.com/github/Blane187/RVC-notebook/blob/main/Easy_GUI.ipynb

[feral halo]

Ayo? @honest mantle level 1 !!!

[honest mantle]

Could you try?

Tell me if it works

[versed jetty]

hi iam cookie monster

[narrow badge]

hello everyone I am a music producer and would love to create with Ai

[honest mantle]

Hi

[fleet ingot]

hello

[honest mantle]

Where are u from

[manic kindle]

hello

[native garden]

Hi

[elder willow]

Can someone fix that AI cover gen collab

@wintry sinew can someone here can fix that

[vale needle]

Hello Guys. who knows how you can hack your own lost google account

[honest mantle]

I guess you must have a recovery emai adress

[vale needle]

Так получилось что этот аккаунт старый, год как забыл пароль, а там мои фотки и так далее

Oh sorry

It so happened that this account is old, I forgot my password a year ago, and there are my photos and so o
n

[ocean osprey]

Hello everyone, I would like to hear some advice from you. Please take care of me.

[covert lagoon]

hi

[honest mantle]

Search google recovery

[storm meadow]

If you have access to a recovery email or phone number you might be able to recover it

[lusty island]

Hello! i'm new here, how are you going?

and btw I'm gonna testing voice changer ^^

[autumn tusk]

how do u reduce the latency in the voice changer

[vale needle]

Google says that such an account does not exist

[versed jetty]

iam in the night because iam from egypt

[stable hazel]

@stark scarab how can i make ai covers on my browser send me google colab link

[stark scarab]

-colab

[rare sorrel]

Suggestions for @stable hazel

☁️ Google Colabs

• Applio, by IA Hispano Google Colab
• AICoverGen-WebUI, modded by Hina Google Colab
• AICoverGen-NoWebUI [English], by Ardha, fixed by Eddy, Hina and Gdr Google Colab
• AICoverGen-NoWebUI [Spanish], credits to Eddy, Hina and Gdr for translating and fixing Google Colab
• Ilaria RVC, by thestingerx Google Colab
• RVC Disconnected, by Kit Lemonfoot Google Colab
• easyGUI, by rejects Google Colab
• 🆕 UVR5 NO UI for Google Colab, by Eddy Google Colab

How to use?

🤗 Hugginface Spaces

• Applio, by IA Hispano Huggingface Spaces
• Ilaria RVC, by thestingerx Hugginface Spaces
• RVC-HFv2, by r3gm Huggingface Spaces
• AICoverGen, by r3gm Huggingface Spaces
• Advanced RVC Inference, by r3gm Huggingface Spaces
• RVC v2 Huggingface version, by Clebersla Huggingface Spaces

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

[stark scarab]

This is more specific for AI Covers:
https://colab.research.google.com/drive/1u1brjK8IZt647UsbZuGYfW29oFM2I4tk?usp=sharing

Inside u can find a guide

[versed jetty]

any one can speak arabic here

any one is muslim here

[stable hazel]

i am muslim but can't speak arabic

[half basalt]

This whole audio rerouting without VAC is a lot more complicated than I thought.

[versed jetty]

ok thats very good man

[limber haven]

Hi guys, I need your help. I have 439 voice samples of a specific actor, and about 50,000 voice lines scrambled in a folder, with different actors. Can I search for this specific voice actors lines by creating a voice model with his samples?

[versed jetty]

you read quran

or no

[stable hazel]

yes

[feral halo]

Ayo? @stable hazel level 1 !!!

[stable hazel]

i can read

but can't speak or understand

[versed jetty]

why you cant understand

[stable hazel]

because i am not arabic

[versed jetty]

ok i can learn you if you want its your choice

[stable hazel]

na it's fine

[versed jetty]

man

[feral halo]

Ayo? @versed jetty level 5 !!!

[versed jetty]

the god is greatest

iam arabic muslim my name is mohamed

[wind stratus]

hi mohamed

[versed jetty]

hi are you muslim too

[storm meadow]

no

[wind stratus]

sorry im not and cant speak arabic either

[limber haven]

I have 439 voice samples of a specific actor, and about 50,000 voice lines scrambled in a folder, with different actors. Can I search for this specific voice actors lines by creating a voice model with his samples?

[rare sorrel]

🎙️ Realtime Voice Changer (Local)

Guide for W-Okada's RealTimeVoiceChangerClient

Description: This one has more options and is somewhat more convenient to use than Real-time RVC.
Link: Rentry
Credits: ravencutie21

How to use real time conversion with RVC (voice changer)

Description: More accurate than W-Okada's, but requires a decent GPU.
Link: Google Docs
Credits: teas.hi

RVC TTS

Description: Text to speech with RVC
Link: Rentry

[stable hazel]

is there is any local software ? when i am trying these non of them working from me

[stark scarab]

-local

[rare sorrel]

Suggestions for @stable hazel

🖥️ 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

Need some help? 🤔

You can find more info on the #1159513888199540817 channel. If you can't find your answer, feel free to ask for help in #✨│ai-help. Credits to Faze Masta and Antasma for compiling these links.

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

[minor blade]

Eddy te mandé DM

Quiero que me ayudes en algo.

[vivid dagger]

hello

[quartz talon]

my voice changer sounds really deformed anyone know how to fix it?

[feral halo]

Ayo? @quartz talon level 1 !!!

[forest quarry]

can you give an example of how it sounds like? you can record audios in the gui

[quartz talon]

sure

[dull hare]

train

[quartz talon]

sent you a vid of what it sounds like

[wanton path]

hi

https://tenor.com/view/hello-hi-friend-penguin-waving-gif-17013804

I'm new to the server

[forest quarry]

ello

[wanton path]

so I do not know much abt the server yet

[knotty anvil]

ello

[forest quarry]

ello

https://media.discordapp.net/attachments/1159290139609137264/1182460305163427982/ezgif-3-9a2dedb470.gif?ex=6584c6f9&is=657251f9&hm=68fbace0527182ce8f089bbd1fb821e1b73549a9a7ae3ddcc2411fc2a0ff8185&

[glossy bloom]

Yo

[glacial hull]

Does anyone here know how to make this? https://www.youtube.com/watch?v=Y3ok-96DolA

[feral halo]

Ayo? @glacial hull level 2 !!!

[signal cairn]

-rvc

[rare sorrel]

Documentation

📚 All-In-One English documentation

https://docs.aihub.wtf/

❔ English FAQ

https://docs.aihub.wtf/faq

Credits: AI HUB FRANCE

✨ More guides

Full AI Voice Model Training Guide (Local)

Link: YouTube
Credits: Christopher Villanueva

Model training with Mainline RVC

Link: Rentry
credits: Raven (ravencutie21)

AICoverGen Colab Guide

Link: Google Docs
Credits: Eddy (Spanish Helper)

Create a model with RVC disconnected (colab)

Link: Google Docs
Credits: Angetyde

How To Make an AI Cover With Ilaria RVC

Link: Rentry
Credits: 👽 Julia (ailen2091)

[tough frigate]

hello what app do yall voice for real time voice changing with models?

[mighty oyster]

@signal cairn está ocorrendo algum problema no arquivo do Colab para fazer os covers? O que estou utilizando é o Easy GUI (for RVC v2).

Sempre recebo o mesmo erro quando executo a célula para baixar o arquivo .zip de meu Drive.

[forest quarry]

https://rentry.co/VoiceChangerGuide

[tough frigate]

thanks is there a video on how to set it up ?

[signal cairn]

colab antigo

-colab

[forest quarry]

most videos have misinfo, so i wouldnt recommend em

[rare sorrel]

☁️ Google Colabs

• Applio, by IA Hispano Google Colab
• AICoverGen-WebUI, modded by Hina Google Colab
• AICoverGen-NoWebUI [English], by Ardha, fixed by Eddy, Hina and Gdr Google Colab
• AICoverGen-NoWebUI [Spanish], credits to Eddy, Hina and Gdr for translating and fixing Google Colab
• Ilaria RVC, by thestingerx Google Colab
• RVC Disconnected, by Kit Lemonfoot Google Colab
• easyGUI, by rejects Google Colab
• 🆕 UVR5 NO UI for Google Colab, by Eddy Google Colab

How to use?

🤗 Hugginface Spaces

• Applio, by IA Hispano Huggingface Spaces
• Ilaria RVC, by thestingerx Hugginface Spaces
• RVC-HFv2, by r3gm Huggingface Spaces
• AICoverGen, by r3gm Huggingface Spaces
• Advanced RVC Inference, by r3gm Huggingface Spaces
• RVC v2 Huggingface version, by Clebersla Huggingface Spaces

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

[forest quarry]

so yeah, written guide ftw

[tough frigate]

alrighty ill try to follow the tut on the site thanks

[forest quarry]

yw :)

[mighty oyster]

Qual desses que é o atualmente utilizado? Não é para treinamento, e sim para converter diretamente o áudio.

[feral halo]

Ayo? @mighty oyster level 1 !!!

[rare sorrel]

🎙️ Realtime Voice Changer (Local)

Guide for W-Okada's RealTimeVoiceChangerClient

Description: This one has more options and is somewhat more convenient to use than Real-time RVC.
Link: Rentry
Credits: ravencutie21

How to use real time conversion with RVC (voice changer)

Description: More accurate than W-Okada's, but requires a decent GPU.
Link: Google Docs
Credits: teas.hi

RVC TTS

Description: Text to speech with RVC
Link: Rentry

[tough frigate]

@forest quarrydis shi bit confusing wheres the tool or thing i download 😂 if theres any vid u can send that will be better

[feral halo]

Ayo? @tough frigate level 1 !!!

[forest quarry]

it links to what you will download

just read carefully, take ur time

[tough frigate]

i guess ill set it up later i thought it was gonna be fast

cause i was bout to do some overwatch voice trollign

[wintry sinew]

hina made that colab

[feral halo]

Ayo? @wintry sinew level 160 !!!

[boreal locust]

салам

что делать если в voicechanger белый экран без ничего

[rare sorrel]

Suggestion for @sage musk

Documentation

📚 All-In-One English documentation

https://docs.aihub.wtf/

❔ English FAQ

https://docs.aihub.wtf/faq

Credits: AI HUB FRANCE

✨ More guides

Full AI Voice Model Training Guide (Local)

Link: YouTube
Credits: Christopher Villanueva

Model training with Mainline RVC

Link: Rentry
credits: Raven (ravencutie21)

AICoverGen Colab Guide

Link: Google Docs
Credits: Eddy (Spanish Helper)

Create a model with RVC disconnected (colab)

Link: Google Docs
Credits: Angetyde

How To Make an AI Cover With Ilaria RVC

Link: Rentry
Credits: 👽 Julia (ailen2091)

[quartz talon]

anyone know how to fix static sound on the voice changer?

[cosmic helm]

How do I use the guides command?

[covert lake]

ask in #🔍│help-w-okada / #1192011222023950368

[covert lake]

-guides

[rare sorrel]

Suggestion for @cosmic helm

Documentation

📚 All-In-One English documentation

https://docs.aihub.wtf/

❔ English FAQ

https://docs.aihub.wtf/faq

Credits: AI HUB FRANCE

✨ More guides

Full AI Voice Model Training Guide (Local)

Link: YouTube
Credits: Christopher Villanueva

Model training with Mainline RVC

Link: Rentry
credits: Raven (ravencutie21)

AICoverGen Colab Guide

Link: Google Docs
Credits: Eddy (Spanish Helper)

Create a model with RVC disconnected (colab)

Link: Google Docs
Credits: Angetyde

How To Make an AI Cover With Ilaria RVC

Link: Rentry
Credits: 👽 Julia (ailen2091)

[cosmic helm]

Tyyy

[versed jetty]

................

.......................................................................................................................................................................

[gusty dagger]

a

[snow sedge]

Damn that's crazy how fast voice changer can actually run on P100 or T4...

[forest quarry]

How'd you get it running that fast? Kaggle? In Colab I got 400ms if I were lucky

[snow sedge]

I use kaggle, yeah. Though the total latency heavily depends on the machine that I get

I was very lucky to get a machine that had only 50ms ping a few times, so total res was around 110ms with these settings

[timid musk]

ai hub users exploiting cloud gpu’s until the day they die

[wintry sinew]

smh

[quiet cargo]

Has someone made a website that shows all the voice models. It's so hard to find voice models on discord

Or is anyone working on it

[timid musk]

https://weights.gg

we get new voice models before weights as they get most from us anyway just FYI

[cloud mirage]

helo

[long raven]

need help

[amber light]

with what

[cosmic helm]

Anyone know what voice model duckus uses?

[amber light]

specifically

[long raven]

@amber light

[amber light]

???

[long raven]

i need help with the config

i alr got the model but the config

[amber light]

voice changer?

[long raven]

yes

caseoh

[scenic imp]

someone maybe can help me find a realistic voice for men?

[amber light]

yeah try asking in #🔍│help-w-okada

[scenic imp]

like I just need something that sounds realistic it doesn't matter if its a popular man or not

[amber light]

have you looked through the lists

[zealous yew]

no

[amber light]

aint been in ai hub long enough to know this is the norm

[forest quarry]

Don't think here is the best place to ask

A quick YT search should work ig

[elder willow]

Oke

[rare sorrel]

Documentation

📚 All-In-One English documentation

https://docs.aihub.wtf/

❔ English FAQ

https://docs.aihub.wtf/faq

Credits: AI HUB FRANCE

✨ More guides

Full AI Voice Model Training Guide (Local)

Link: YouTube
Credits: Christopher Villanueva

Model training with Mainline RVC

Link: Rentry
credits: Raven (ravencutie21)

AICoverGen Colab Guide

Link: Google Docs
Credits: Eddy (Spanish Helper)

Create a model with RVC disconnected (colab)

Link: Google Docs
Credits: Angetyde

How To Make an AI Cover With Ilaria RVC

Link: Rentry
Credits: 👽 Julia (ailen2091)

[normal jungle]

Any room we can request songs with the voice we want?

[limber ermine]

oops'

[violet igloo]

~

[storm meadow]

No only models

If you want a cover do it yourself #1159513888199540817 #✨│ai-help

[harsh lily]

Hi guys! Anyone would have Snoop dogg model?

[storm meadow]

#1175430844685484042

[bronze citrus]

whats the link to the rvc

or google collab thing i mean

[forest quarry]

https://docs.aihub.wtf/ has links to colabs and guides for em

[normal jungle]

can i toss someone a ocuple bucks to do it for me? i heard stan by eminem with simpsons voice on youtube can can't seem to find it ever since.

[amber light]

you can learn how to make one

see the doc mj sent above

[bitter rock]

this is the first time i ever chatted here

[covert lake]

Welcome

[sinful mist]

Does anyone voice Ivan Cornejo?

[pale wing]

Such a cool resource here

[ember grotto]

yo

[amber light]

#1175430844685484042 or weights.gg

[sinful mist]

You can pass it on to me

[feral halo]

Ayo? @sinful mist level 1 !!!

[cosmic helm]

every model i import sounds like ass

the base 4 work pretty well idk why

[broken narwhal]

can i get the creator role if i had it on the old discord

proof is searchable on weights.gg

https://weights.gg/models/clm739cqp1wpacctc8f1akfft

[bleak river]

Do you know how to remove accent from a model?

[feral halo]

Ayo? @bleak river level 1 !!!

[mossy mulch]

hi so like

[white shale]

Hello, I would like help, how can I make covers with the voices?

[mossy mulch]

I need help with making my voice dataset

[storm meadow]

-guides

[rare sorrel]

Suggestion for @white shale

Documentation

📚 All-In-One English documentation

https://docs.aihub.wtf/

❔ English FAQ

https://docs.aihub.wtf/faq

Credits: AI HUB FRANCE

✨ More guides

Full AI Voice Model Training Guide (Local)

Link: YouTube
Credits: Christopher Villanueva

Model training with Mainline RVC

Link: Rentry
credits: Raven (ravencutie21)

AICoverGen Colab Guide

Link: Google Docs
Credits: Eddy (Spanish Helper)

Create a model with RVC disconnected (colab)

Link: Google Docs
Credits: Angetyde

How To Make an AI Cover With Ilaria RVC

Link: Rentry
Credits: 👽 Julia (ailen2091)

[bleak river]

Hello, do you know how to remove accent from a model(English to Portuguese)?

[storm meadow]

Don’t use the index

[bleak river]

I still has the accent

*it

[tender depot]

hello

[bleak river]

Are you saying to not train the index?

[storm meadow]

no

The index acts like a accent

[violet igloo]

[storm meadow]

Not having it makes no accent

Unless it’s baked in

[bleak river]

It's baked in

[bleak river]

Well, when I try to use the model it succeed to create the audio but it says that index file was not used

[bleak river]

And still has the accent (American accent)

[storm meadow]

Is it supposed to have it

Do you have a American accent

[violet igloo]

[bleak river]

I'm trying to translate an english video in Portuguese(I'm fluent in Portuguese) but it has the American accent when I speak Portuguese.

[mossy mulch]

how the hell do I train a voice?

[feral halo]

Ayo? @mossy mulch level 1 !!!

[mossy mulch]

without using a dumb colab

[storm meadow]

Try undertraining the model (if you made the model) if removing the index doesn’t work

[mossy mulch]

or local software

[storm meadow]

Without colab and local software?

Or via local software without colab

[mossy mulch]

I meant like a easy gui

because I have audio files

[storm meadow]

so local option?

[mossy mulch]

ok

[storm meadow]

What’s your GPU

[bleak river]

https://youtu.be/Evs8Or2Qy9A?si=WZlhPQikqER36EZ-

Using Google colab

[storm meadow]

Doesn’t have a gui

And I have never heard of that colab

[bleak river]

I train that way

[storm meadow]

-colab Applio has a UI and training

[rare sorrel]

☁️ Google Colabs

• Applio, by IA Hispano Google Colab
• AICoverGen-WebUI, modded by Hina Google Colab
• AICoverGen-NoWebUI [English], by Ardha, fixed by Eddy, Hina and Gdr Google Colab
• AICoverGen-NoWebUI [Spanish], credits to Eddy, Hina and Gdr for translating and fixing Google Colab
• Ilaria RVC, by thestingerx Google Colab
• RVC Disconnected, by Kit Lemonfoot Google Colab
• easyGUI, by rejects Google Colab
• 🆕 UVR5 NO UI for Google Colab, by Eddy Google Colab

How to use?

🤗 Hugginface Spaces

• Applio, by IA Hispano Huggingface Spaces
• Ilaria RVC, by thestingerx Hugginface Spaces
• RVC-HFv2, by r3gm Huggingface Spaces
• AICoverGen, by r3gm Huggingface Spaces
• Advanced RVC Inference, by r3gm Huggingface Spaces
• RVC v2 Huggingface version, by Clebersla Huggingface Spaces

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

[bleak river]

There is the RVC disconnected

[storm meadow]

-local for local the best choice is still orginal RVC

[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

Need some help? 🤔

You can find more info on the #1159513888199540817 channel. If you can't find your answer, feel free to ask for help in #✨│ai-help. Credits to Faze Masta and Antasma for compiling these links.

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

[storm meadow]

I know

[violet igloo]

🫂

[mossy mulch]

it is a pain

[storm meadow]

That’s why I’m recommending you other options

[bleak river]

Do you know how to remove accent?

[feral halo]

Ayo? @bleak river level 2 !!!

[arctic galleon]

hi i'm new to ai and was looking for some free resources

any ideas

[bleak river]

Google colab

[storm meadow]

what are you planning to use ai for

We mainly do ai music/covers here but certain can help with ai art and stuff

[bleak river]

About the accent, I know that creating audios(using as inference )through elevenlabs, the accent is really reduced but it kind sounds like robotic, I think that if used that audio from Elevenlabs(Portuguese audio) to train a new model in Portuguese would work but I'm afraid that could sound robotic.

What do you think?

[bleak river]

What your opinion about what I said?

[amber light]

0 ratio index

basically just turning it off

[bleak river]

I'm using hugging face for inference

How to do that?

[amber light]

theres supposed to be a slider saying index ratio

you turn it down to 0

[bleak river]

There's search feature ratio

But it doesn't has index ratio

[amber light]

yes

think that should be it

[bleak river]

Thank you

[empty adder]

hi

[bleak river]

Unfortunately it still has the accent

[amber light]

hmm

wait then

[mellow cloud]

hi @steel quiver

[amber light]

yeah there should be an index ratio slider

[mellow cloud]

hey @amber light can u do something for me

[amber light]

something similar to this

[amber light]

depends

what is it

[mellow cloud]

would u create me a model of Silver Spoon from Inanimate Insanity

[amber light]

#1159289738314919936

[bleak river]

For me it doesn't have this thing

[bleak river]

I'm using hugging face inference mangio-RvC fork hf

[amber light]

-colab

[rare sorrel]

☁️ Google Colabs

• Applio, by IA Hispano Google Colab
• AICoverGen-WebUI, modded by Hina Google Colab
• AICoverGen-NoWebUI [English], by Ardha, fixed by Eddy, Hina and Gdr Google Colab
• AICoverGen-NoWebUI [Spanish], credits to Eddy, Hina and Gdr for translating and fixing Google Colab
• Ilaria RVC, by thestingerx Google Colab
• RVC Disconnected, by Kit Lemonfoot Google Colab
• easyGUI, by rejects Google Colab
• 🆕 UVR5 NO UI for Google Colab, by Eddy Google Colab

How to use?

🤗 Hugginface Spaces

• Applio, by IA Hispano Huggingface Spaces
• Ilaria RVC, by thestingerx Hugginface Spaces
• RVC-HFv2, by r3gm Huggingface Spaces
• AICoverGen, by r3gm Huggingface Spaces
• Advanced RVC Inference, by r3gm Huggingface Spaces
• RVC v2 Huggingface version, by Clebersla Huggingface Spaces

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

[amber light]

use a newer version

or newer colabs/spaces

[timid musk]

ski

[amber light]

that ones outdated

yes

[timid musk]

no jackst

[amber light]

tea

[timid musk]

no jacket

[amber light]

Are you still looking

Buy the one that looks the closest to it

[bleak river]

Unfortunately I don't know how to upgrade

[amber light]

like i said

theres colabs/spaces up there where automaze sent

you can use those

[amber light]

this one is old

[near mountain]

Guys with music models like rap is it bad to keep the intro before they rap and outro when they talk cause it might confuse the model or hurt it

[sonic tundra]

hi

[thorny cedar]

guys

why does my w okada voice changer have no crepe

its only rmvpe and crepe_tiny

[tepid basin]

Yeah, generally I cut that out of my dataset. Usually if you want talking samples, it's better to just rip it from an interview or something.

People create datasets with 50/50 speaking and singing, they sound decent, wont hurt to try to mix the data. But usually if I'm training a rapper model, I make it 100% rap/singing

Stuff like Travis Scott's Rodeo skits come to mind; id cut those for sure

One more thing; not sure what rapper you're training on, but if it's a huge rapper, there might be some leaked stems floating around, usually on Spreadsheets. I cant help you find them sadly

[tepid basin]

i mean you should really only use RMVPE anyway. idk why they're missing tho

[teal ivy]

thankyou thats vbery useful info i appreciate it

[feral halo]

Ayo? @teal ivy level 9 !!!

[tepid basin]

np :)

[teal ivy]

and 1 last thing do u still need to cut the audio into seconds of 5 or 10 or just a whole audio file will do now

[elder willow]

hi

[coarse vessel]

in audacity, Analyze -> Label sounds, then export as multiple files

[hardy kestrel]

https://tenor.com/view/cat-cats-explosion-explodes-cat-explodes-gif-10131775219258044191

[spice lark]

can i use mp3/wav files in wordband instead of a youtube link

or do i gotta upload it to yt

[storm meadow]

Try Astra labs instead if you want to use a file

[spice lark]

oki

[vivid wasp]

Does anyone have a link to Gerard Way's voice? for My Chemical Romance

[feral halo]

Ayo? @vivid wasp level 1 !!!

[radiant verge]

👋

[storm meadow]

Check weights gg

You should be okay (I promise)

[still crane]

whats epochs ??

[feral halo]

Ayo? @still crane level 1 !!!

[storm meadow]

Training cycles, more or less does not mean better

You can ignore the number

[still crane]

Ty

And whats rvc v2 ?

Or Ova2 ?

[storm meadow]

Model architecture, all new versions of RVC have support for it

[storm meadow]

That’s a pretrain for models to train off of

[vivid wasp]

thanks

[still crane]

Dose rvc and okada have the same goal ?

And how can i know whats better for me ?

[coarse vessel]

non-realtime vs realtime

[storm meadow]

w-Okada is real time, RVC is mainly for covers

[empty sapphire]

yo whats the best collab notebook for RVC now

im getting errors on Easy Gui (for RVC v2)

[storm meadow]

-colab

[rare sorrel]

Suggestions for @empty sapphire

☁️ Google Colabs

• Applio, by IA Hispano Google Colab
• AICoverGen-WebUI, modded by Hina Google Colab
• AICoverGen-NoWebUI [English], by Ardha, fixed by Eddy, Hina and Gdr Google Colab
• AICoverGen-NoWebUI [Spanish], credits to Eddy, Hina and Gdr for translating and fixing Google Colab
• Ilaria RVC, by thestingerx Google Colab
• RVC Disconnected, by Kit Lemonfoot Google Colab
• easyGUI, by rejects Google Colab
• 🆕 UVR5 NO UI for Google Colab, by Eddy Google Colab

How to use?

🤗 Hugginface Spaces

• Applio, by IA Hispano Huggingface Spaces
• Ilaria RVC, by thestingerx Hugginface Spaces
• RVC-HFv2, by r3gm Huggingface Spaces
• AICoverGen, by r3gm Huggingface Spaces
• Advanced RVC Inference, by r3gm Huggingface Spaces
• RVC v2 Huggingface version, by Clebersla Huggingface Spaces

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

[fathom patio]

Why does it not work when I load my model

[zenith hamlet]

Is there any issues with colabs that have UI?

[trail coral]

google bans them

[zenith hamlet]

I meant at the moment

I know google bans them

[trail coral]

all i can say is use it with risk

[zenith hamlet]

I already risked enough by making over 60 ai covers with 2-3 mixing effort lol,i don't fear getting banned

[buoyant forge]

hola

[maiden shadow]

hi

[elder willow]

Guys help, my Okada voice changer download is stuck at 1%

[feral halo]

Ayo? @elder willow level 2 !!!

[faint warren]

apologies if its a dumb question, but the number of epochs refers to how well trained the model is?

[abstract compass]

does anyone know any free voice cloning website

[bold saffron]

nah that's a myth

model can be under train or over train if you just train it for the no. of epoch count

[faint warren]

ah okay thank you! just got into it today so i wasnt sure what they were

[feral halo]

Ayo? @faint warren level 1 !!!

[bold saffron]

no worries, a geniune doubt i guess

[faint warren]

is there a good range to strive for when looking for models

[bold saffron]

anything equal to or more than 200/300 is claim to be good

well all epoch count thing fails if the model's trained on a bad dataset

[faint warren]

so a model claiming to be for an uwu girl trained only on uwus probably isnt a good set

[bold saffron]

uwu's lol

[coarse vessel]

meme model, even there is mc villager model

[faint warren]

i saw it on the applio site LOL

[bold saffron]

if anyone's looking for uwu girl type voice model just look here https://discord.com/channels/1159260121998827560/1175430844685484042

or pay to make your custom model

[bright arch]

what model are you looking for

[faint warren]

right now i was just browsing for some interesting models and seeing what caught my attention

[bright arch]

oh you can check them out either in #1175430844685484042 or weights.gg

[faint warren]

Thank you! ill def take a look

[covert urchin]

Hello, does anyone know where I can install the RVC program to use AI?

[lean blaze]

.wk

[indigo pike]

xaviersobased in AI HUB

👑  ⁦Adiiose⁩ - 4206 plays
 2.  ⁦remy⁩ - 3988 plays
 3.  ⁦exan⁩ - 3141 plays
 4.  ⁦luminal⁩ - 2823 plays
 5.  ⁦yeresin⁩ - 2299 plays
 6.  ⁦uh1c⁩ - 993 plays
 7.  ⁦sqwab⁩ - 962 plays
 8.  ⁦mox⁩ - 862 plays
 9.  ⁦gtemq⁩ - 767 plays
 10. ⁦osc⁩ - 765 plays
** 11.**  ⁦gilbert⁩ - 725 plays
 12.  ⁦buga⁩ - 675 plays
 13.  ⁦cxvz⁩ - 615 plays
 14.  ⁦Frankie⁩ - 603 plays

cloud rap - dark plugg - sigilkore Artist - 295 listeners - 40588 plays - 137 avg

[flint musk]

hi

[minor glen]

привет

[sacred oasis]

I'm saif

[hollow jetty]

what is epoch?

[magic cave]

is there any other wbsite other than Google Collab where I can convert songs into AI covers, google collab isnt working for me for some reason, it wont convert Gordon Ramsay

[dense harness]

what is the best realtime VC for AMD?

[elder willow]

-colab

[rare sorrel]

☁️ Google Colabs

• Applio, by IA Hispano Google Colab
• AICoverGen-WebUI, modded by Hina Google Colab
• AICoverGen-NoWebUI [English], by Ardha, fixed by Eddy, Hina and Gdr Google Colab
• AICoverGen-NoWebUI [Spanish], credits to Eddy, Hina and Gdr for translating and fixing Google Colab
• Ilaria RVC, by thestingerx Google Colab
• RVC Disconnected, by Kit Lemonfoot Google Colab
• easyGUI, by rejects Google Colab
• 🆕 UVR5 NO UI for Google Colab, by Eddy Google Colab

How to use?

🤗 Hugginface Spaces

• Applio, by IA Hispano Huggingface Spaces
• Ilaria RVC, by thestingerx Hugginface Spaces
• RVC-HFv2, by r3gm Huggingface Spaces
• AICoverGen, by r3gm Huggingface Spaces
• Advanced RVC Inference, by r3gm Huggingface Spaces
• RVC v2 Huggingface version, by Clebersla Huggingface Spaces

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

[elder willow]

-colab

[rare sorrel]

☁️ Google Colabs

• Applio, by IA Hispano Google Colab
• AICoverGen-WebUI, modded by Hina Google Colab
• AICoverGen-NoWebUI [English], by Ardha, fixed by Eddy, Hina and Gdr Google Colab
• AICoverGen-NoWebUI [Spanish], credits to Eddy, Hina and Gdr for translating and fixing Google Colab
• Ilaria RVC, by thestingerx Google Colab
• RVC Disconnected, by Kit Lemonfoot Google Colab
• easyGUI, by rejects Google Colab
• 🆕 UVR5 NO UI for Google Colab, by Eddy Google Colab

How to use?

🤗 Hugginface Spaces

• Applio, by IA Hispano Huggingface Spaces
• Ilaria RVC, by thestingerx Hugginface Spaces
• RVC-HFv2, by r3gm Huggingface Spaces
• AICoverGen, by r3gm Huggingface Spaces
• Advanced RVC Inference, by r3gm Huggingface Spaces
• RVC v2 Huggingface version, by Clebersla Huggingface Spaces

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

[glacial hull]

But RVC doesn't really work here because of voice expressions

Idk how to explain what I mean

[stable hazel]

-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

Need some help? 🤔

You can find more info on the #1159513888199540817 channel. If you can't find your answer, feel free to ask for help in #✨│ai-help. Credits to Faze Masta and Antasma for compiling these links.

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

[solemn flint]

Have you found an editor?

its Too Amazing !

https://youtu.be/U9l9wZ1Hjfo?Sub_Confirmation=1

[zenith hamlet]

Voicemy.ai and fakeyou.com

[fallen ruin]

https://cdn.discordapp.com/attachments/1223798464622760033/1233921730238484480/cat-silly.mp4?ex=662edaa9&is=662d8929&hm=9f7d79ecd77780180b850a6339dfd716ea6e789a3bee1f7db192df3a3164e1a7&

[feral halo]

Ayo? @opal fable level 1 !!!

[static ledge]

https://huggingface.co/blaise-tk/TITAN/tree/main/models/medium/40k/model

is model here initialized with random weights?

[naive sonnet]

sorry for the trouble but listen to this with subtitles https://youtu.be/zhEiduBxX4c?si=9nZGW_cwWIFenxyT

Official

[elder willow]

[versed spindle]

hello everyone

[magic cave]

Thank you so much 😄

[feral halo]

Ayo? @magic cave level 2 !!!

[sour remnant]

is anyone facing issues with ilaria rvc? i cant convert audio files as it gives me an "ERROR"

[zenith hamlet]

Did you try closing and reopening the tab

[feral halo]

Ayo? @zenith hamlet level 9 !!!

[sour remnant]

yea i did

is there any other collab like ilaria to convert audio using ai?

[zenith hamlet]

Maybe applio but it's slow asf tbh, there's also no ui colabs but i prefer with ui

They also add weirdass settings reverb and etc

[sour remnant]

oh

it was working fine until yesterday

giving errors once i convert audio on the gradio page/link whatever

also, running the gradio cell gives me an AttributeError in the end

AttributeError: 'NoneType' object has no attribute 'tobytes'

and it doesnt stop running

any fix?

also recently it has started asking when i run the start cell that it detects no GPU and i just conitnue

[earnest kite]

Is there any way to use RVC V2 models with Synthesizer V or Vocaloid? I really wanna be able to use custom trained models with these software

[indigo anchor]

are you asking if you want to use them as inference audio?

or like... training a model off of them?

[sour vigil]

nop

[indigo anchor]

well you can use it for inference audio. just double check the terms of service for whatever software you're trying that with

but most times out of ten you won't be able to train a model off of them

[sour vigil]

they probably meant somehow using an rvc model as a voicebank in any vocaloid app

[indigo anchor]

oh... idk how that'd work

[sour vigil]

it wouldnt lol

[indigo anchor]

closest we have to that is diffsinger, but that requires you labeling the data

which takes hours

[elder willow]

anyone got a travis scott utopia era model

mine aint soundin great rn

[indigo anchor]

you asked him if you could make a model off his voice?

idk why I even bothered asking that

[feral halo]

Ayo? @indigo anchor level 1 !!!

[indigo anchor]

that's like asking someone who commits tax fraud on the daily if they paid their taxes

[quartz roost]

That's kind of what this does. https://youtu.be/k32AwsAbc6A?si=1YE2BEjPIOtgOXjE

[indigo anchor]

not really. vocoflex is just a voice changer plugin. you can load synthv output into it but that's it

[quartz roost]

Yeah you mentioned diffsinger already

[indigo anchor]

Maybe dt was over cautious with this. I don't think this would fool anyone

[reef idol]

hello

[errant sphinx]

https://tenor.com/view/okokokok-clapping-hands-yay-excited-gif-24387954

[mental jasper]

are there any good paid ai voice changers

[quartz roost]

There is paid professional studio ai voices.

[mental jasper]

where do i find then

[feral halo]

Ayo? @mental jasper level 1 !!!

[quartz roost]

What would you like to make, there is several

[low mason]

turkish models why?

[limber haven]

live or for recordings?

[feral halo]

Ayo? @limber haven level 1 !!!

[low mason]

How can I find Müslüm Gürses voice model?

[indigo anchor]

who?

[low mason]

I urgently need audio file of Turks

[feral halo]

Ayo? @low mason level 1 !!!

[low mason]

How do I level up? At what level does it open?

[dense harness]

what's the best realtime rvc client for AMD GPUs?

[pliant solstice]

Learn to create AI sales bot for your ecommerce business: https://www.youtube.com/watch?v=KNKKgxDMIZc

[feral halo]

Ayo? @pliant solstice level 1 !!!

[mental jasper]

live

[bold saffron]

W-Okada

[dense harness]

too much delay required

[bold saffron]

https://rentry.co/VoiceChangerGuide

[dense harness]

like 200-300

and my gpu is top tier

[bold saffron]

follow this guide thorougly

[indigo anchor]

doesn't mangio have a realtime client built in?

[mental jasper]

guys why do when i use okada and i hear weird noises when i use it like when i breath it makes an alien noise

[indigo anchor]

that's likely because the data doesn't have breath noises in it

[mental jasper]

im scared of those noises😭

[indigo anchor]

just try to be careful not to make those noises

[feral halo]

Ayo? @indigo anchor level 2 !!!

[forest surge]

yo

[worn prism]

Wow so many voices

[feral halo]

Ayo? @worn prism level 1 !!!

[indigo anchor]

there's a go-realtime-gui in mangio's files

is it good for realtime standards? idk

I think it's alright

[errant sphinx]

DARK SYSTEM..!!

Everyone so dark system..!!

[indigo anchor]

what?

[timid musk]

how do you make an ai cover

[timid musk]

yes

i use that over okada personally

there’s a whole guide i wrote on it

-realtime

[rare sorrel]

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

[lilac loom]

Random question: Has anyone ever used GPT-SoVITS to make a Team Fortress 2 video in Japanese?
https://huggingface.co/spaces/kevinwang676/gpt_sovits_jay

[errant sphinx]

Oh you new here Hai..!!

[indigo anchor]

so, personally I'd do this

1. pick the song you want to cover
2. sing over the song yourself or use something like synthv to render it (using stem-isolated vocals won't give you the cleanest results)
   2a. make sure the tos of the synth your using allows you to do this if you choose to use a vsynth input
3. load the audio of the vocals into the model of your choice

[bold saffron]

ayo bro got it

[errant sphinx]

So, what is about it..?! 🤔

[indigo anchor]

also this goes without saying, but it's recommended to use models you actually have the rights to. while making a cover of a song with 2pac or travis scott may be fun, you didn't get the rights to use their voice

[timid musk]

i thank you but i was memeing

[errant sphinx]

Yeah, TRUE WAS..................

[timid musk]

i have been doing this for a year 😭🤣

[indigo anchor]

ah

thanks for letting me know

[north jewel]

Hello, how do you guys make ai song cover I'm new😓

[indigo anchor]

#🧬│ai-chat message

[timid musk]

LOL

[worn prism]

Well there’s a lot of videos on how to do it

[north jewel]

Ok thanks

[timid musk]

at ai hub we don’t care about TOS on other platforms 😎

[worn prism]

Well for me I use rvc gui.

[elder willow]

holy fuck i done it

[north jewel]

I tried in some website but there's no ai voice of others character like geto

[elder willow]

i've actually made a voice model

[indigo anchor]

plus, how would you be able to train models off of them anyway? stem isolated recordings? that'd get you a lot of artifacting since the voices are gonna be processed to hell and back for the songs

unless someone here somehow has access to the master recordings

[forest sonnet]

de noise makes a good impression

[feral halo]

Ayo? @forest sonnet level 1 !!!

[forest sonnet]

Noise makes the artifacts from acapella seperation

Pretty sure you can make a good model with the right tools

[elder willow]

Hello guys discord not detecting voice changer at options in input device how i can fix that

[indigo anchor]

do you have vb audio virtual cable installed?

[elder willow]

is that a program ?

i just downloaded voice changer client demo

[indigo anchor]

it's a set of drivers

[elder willow]

probably not

[coarse vessel]

VAC is better for less issues
https://software.muzychenko.net/freeware/vac470lite.zip

[sterile urchin]

you’re practically asking for a C&D at this rate

[indigo anchor]

it does say it's a lite version though

[elder willow]

okay i did download that one

[indigo anchor]

I think it's important to consider that they could potentially, hypothetically, be joking

[sterile urchin]

true

[lament sandal]

FREE NITRO DMS ME

[elder willow]

okay i have download and setup that one in link

[errant sphinx]

https://tenor.com/view/blissful-malevonence-gif-10192619856801101589

[feral halo]

Ayo? @elder willow level 1 !!!

[elder willow]

it shows in input device but cant detect my voice now

virtual audio cable line 1 in my input device

[indigo anchor]

so it says "explicit commercial use is prohibited" in the tos for the trial. it's drivers! what commercial use would be derived from that??

[feral halo]

Ayo? @indigo anchor level 3 !!!

[ruby otter]

ага

[coarse vessel]

yeah no monetization using such AI-generated content

[indigo anchor]

no I'm talking about the drivers themselves

[elder willow]

Does anybody know good voice changer supported rmvpe?

[feral halo]

Ayo? @elder willow level 2 !!!

[coarse vessel]

just install it no need to reboot

[indigo anchor]

mangio has a built-in realtime client

[elder willow]

Oh thanks, Ill try it

[indigo anchor]

I'm just confused on what the tos means. what commercial usage would be derived off of audio drivers? no one is gonna tell you're using those specific ones

[coarse vessel]

either RVC realtime or the one in #🔍│help-w-okada

[indigo anchor]

oh, I've tried the w-okada client before. it worked for a bit, but now it just errors out without giving me an error message

so I pivoted to using the one that's included with mangio

the only issue I have with it is that you have to stop and start the processing every time you want to change something

[elder willow]

I see

[indigo anchor]

for swapping model/index files I can understand that but changing things like the sample length and pitch shift requires you to stop and start as well

[pulsar flume]

God I didnt realise subtitling video with timings was so painful xD

[feral halo]

Ayo? @pulsar flume level 1 !!!

[indigo anchor]

wait are you doing every word individually?

[covert lake]

Hello

[indigo anchor]

Hi

we were just discussing realtime clients

[pulsar flume]

Nah not putting myself through that pain just syncing up the lyrics in subtitles (I did not want to wait 3 hours for it to render again)

Youtubes subititle editor isnt that bad tbh

[indigo anchor]

assuming that the song has singing that's enunciated properly, it shouldn't be too hard

[pulsar flume]

Oh yeah it wasnt hard at all (i've finished now) just time consuming. But hey I rendered the video at 1080p@60 so I wanted to put some effort into it.

1080p@60 because of the visualizer (I wanted it to be smooth)

[pulsar flume]

Realtime clients work great for me just dont sound quite right as a lot of the models are American and not British

[indigo anchor]

well, you can use british models. you just need to speak in a british accent to match

[pulsar flume]

I am british lol

Its the American ones that I cant get right. I cant do a good accent

[indigo anchor]

I understand the sentiment

my friend is british so me trying to use their model produces some funny results
they gave me permission to train a model of them don't worry

[timid musk]

that’s rough

[pulsar flume]

Haha yeah I have a model of my American friend and that just changes the pitch and pretty much nothing else

[timid musk]

index will help with accent

but for realtime it’s kinda odd

[pulsar flume]

See I dont know what index actually does. I change it and I hear no difference whatsoever

[timid musk]

there’s a difference

for realtime it’s not great

and some models don’t even have an index file

[indigo anchor]

what's important with realtime is that your friend isn't gonna have your exact intonation. it'd sound more accurate if you tried to copy that instead of speaking normally

[pulsar flume]

See russian I can kind of do. Its pretty shit but it does sound "ok"

[timid musk]

TOS isn’t a legal document, it’s their personal terms of service

[indigo anchor]

it's best to still follow it anyway. I wouldn't recommend getting on the developers' bad side if they catch you breaking it

[pulsar flume]

EULA is binding but TOS is just what you can and cant do on that platform

[feral halo]

Ayo? @pulsar flume level 2 !!!

[timid musk]

exactly

[pulsar flume]

All you risk is getting banned / blocked

[timid musk]

yup

[pulsar flume]

Theres no legal element

[timid musk]

someone with a brain oh my god it’s been so long

[pulsar flume]

Haha I dont know much about legal stuff but if people take the time to read the first sentence it'll tell you if its binding or not

[indigo anchor]

since I last have seen my son lost to this monster

[indigo anchor]

huh, I just thought they were used interchangeably

I just treat the tos as their end user license agreement at the end of the day

[timid musk]

Tos literally means Terms of Service which means you have to follow those terms in order to use their service

nothing else

[indigo anchor]

so if you break it, they can remove access to the service

best to not risk that in my opinion

[timid musk]

yes

[pulsar flume]

TOS is platform specific and non-binding

EULA is binding

EULA is also platform specific

[indigo anchor]

makes sense, but I do believe it is in the user's best interest to follow both

[timid musk]

yes, eula is that thing you agree to on minecraft not to hack

😭

[indigo anchor]

yet people still choose to skip them entirely

[feral halo]

Ayo? @indigo anchor level 4 !!!

[pulsar flume]

Yup and also includes shit like "dont modify, decrypt or distribute any software" (MCCoderPack be like)

[indigo anchor]

once I get into software development, I'm going to recommend the user to actually sit down and read the thing. maybe have like a questionaire on what you can/can't do or something. if you get all the questions right, you can install your software

[coarse vessel]

basically not to redistribute without permission (publishing as your own)

[sterile urchin]

but you’re still saying you agree and therefore you are bound to them. in quite a few places you are held responsible for what you sign or agree to regardless if you have read them

[feral halo]

Ayo? @sterile urchin level 1 !!!

[sterile urchin]

jell if you do that to me, we are no longer friends

[timid musk]

real, but the whole issue for companies like that is people downloading models of vocolid models or similar didn’t agree to anything

[indigo anchor]

there's probably better ways to go about it however, like making the terms more digestible so it's not 200 pages when it doesn't need to be

[pulsar flume]

Pretty much

[sterile urchin]

true

[coarse vessel]

go free open source

[sterile urchin]

go for gdpr style writing

[upper fern]

.

[hollow thicket]

Does anyone know what they're using to make these remixes? https://www.youtube.com/watch?v=8oDYSJIyilw

[sterile urchin]

short concise and simple language

[pulsar flume]

"Okay now you have read the TOS, recite it exactly"

"You have 5 minutes"

[timid musk]

ppl talking about tos that’s wild

[pulsar flume]

OSS for the win

[indigo anchor]

ok that's probably a little overboard

[sterile urchin]

yeah

[pulsar flume]

But you know they have fully absorbed that information 😉

[frigid condor]

"now that youve read our ToS, wed like you to take 20 mins of ur time to take this quiz"

[indigo anchor]

just have something in place to ensure the user actually read the terms. having a check box for "I agree to the terms" isn't doing anything

[sterile urchin]

most people don’t read the TOS

[indigo anchor]

exactly! that's the problem

[elder willow]

Can I get sound with rx570 ryzen 5 5600x?

[indigo anchor]

there could be something in those terms that's like "you must give a blood sacrifice to the company every odd day" and nobody would bat an eye

[sterile urchin]

as long as you have speakers yes

[coarse vessel]

✅ "By continuing, you agree to the TOS and shit."

[indigo anchor]

since they don't read it

[elder willow]

..

[indigo anchor]

what do you mean by "get sound"?

[elder willow]

voice ai

performance.

[indigo anchor]

is it like... inferencing? training? realtime stuff?

[elder willow]

umm real time?

[sterile urchin]

nope

[elder willow]

ups

Which graphics card can I get good performance with?

[sterile urchin]

you’ll want an nvidia gpu for that so you can use the cuda core

[indigo anchor]

a decently high-performance one from NVIDIA will do the job

[elder willow]

Is 3060 enough?

[timid musk]

okay so number one

[coarse vessel]

it is good

[indigo anchor]

from the last time I was on the server, it seemed so

[timid musk]

okada has a version that works with radeon

so AMD cards work

also

3060 is more than enough

[elder willow]

thankkss

[pulsar flume]

Thats what I have "12gb" and its decent enough. W-OKADA can be used with a chunk size as low as 80 (starts to hiccup here though)

[coarse vessel]

3060 is recommended esp when running with a game

[sour elm]

hey guys, ive been using this guide for ai covers but cant make any voice model work anymore: https://docs.google.com/document/d/1J-JlFGglJY7SpLtC-_3p8lW8CL_WJYqbo5X6CA7OnKM/edit

using Corpse's models from #1175430844685484042 and archives, but none are running in the Colab - giving the* JSONDecodeError: Expecting value: line 1 column 1 (char 0)* error

anyone ran into this issue and knows how to fix, perhaps? also tried downloading the zip file and uploading on own drive to share a link with Colab - no use. has there been a change in ai cover process?

thanks a ton in advance, noob here 🥲

[indigo anchor]

I was able to get realtime clients working on a laptop-grade 1650. idk how I managed to not blow up my laptop this way

[pulsar flume]

#✨│ai-help

[sour elm]

thanks, sorry for wrong ping!

[elder willow]

My graphics card is a little old

[feral halo]

Ayo? @elder willow level 1 !!!

[timid musk]

what colab are you using fuck them

[snow sedge]

They actually work decently. Sometimes I use a laptop with this card as a server

[elder willow]

What do you think would happen if I changed my graphics card to 3060 and my processor was 5600x?

[sour elm]

https://colab.research.google.com/drive/1Gj6UTf2gicndUW_tVheVhTXIIYpFTYc7?usp=sharing%2F#scrollTo=OVQoLQJXS7WX

idk its been working quite well in ~november, but now seems like the models don't run...

[frigid condor]

that would be pretty cool i think

[indigo anchor]

that... seems like a downgrade? idk for sure

[timid musk]

try using any of these

-colab

[rare sorrel]

☁️ Google Colabs

• Applio, by IA Hispano Google Colab
• AICoverGen-WebUI, modded by Hina Google Colab
• AICoverGen-NoWebUI [English], by Ardha, fixed by Eddy, Hina and Gdr Google Colab
• AICoverGen-NoWebUI [Spanish], credits to Eddy, Hina and Gdr for translating and fixing Google Colab
• Ilaria RVC, by thestingerx Google Colab
• RVC Disconnected, by Kit Lemonfoot Google Colab
• easyGUI, by rejects Google Colab
• 🆕 UVR5 NO UI for Google Colab, by Eddy Google Colab

How to use?

🤗 Hugginface Spaces

• Applio, by IA Hispano Huggingface Spaces
• Ilaria RVC, by thestingerx Hugginface Spaces
• RVC-HFv2, by r3gm Huggingface Spaces
• AICoverGen, by r3gm Huggingface Spaces
• Advanced RVC Inference, by r3gm Huggingface Spaces
• RVC v2 Huggingface version, by Clebersla Huggingface Spaces

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

[pulsar flume]

Try RVC-Disconnected (if thats still a thing)

[timid musk]

any of those work

[pulsar flume]

GUI's are known to get blocked by colab faster

[timid musk]

true

[sour elm]

lemme check, thank you so much! the guide was really useful for RVC models, not sure if the reason is in them being old

[timid musk]

all that matters for ai stuff is gpu

[indigo anchor]

first bitcoin miners, now ai cover makers are driving up gpu prices

[sterile urchin]

-dies in Mac- I am so considering swapping to windows

[timid musk]

VRAM specifically

[pulsar flume]

GTP-SoVITS is for TTS right?

[indigo anchor]

yes

[timid musk]

max works still which is surprising

mac

[pulsar flume]

I tried tortoise and it works great but 3 hours of training time is not fun

[timid musk]

one of our sr mods is on a mac lmao

[sterile urchin]

it doesn’t for me it likes to fight me

[indigo anchor]

I've seen people make models for it, but all I've ever seen it work with inferencing wise is 0-shot tts

[timid musk]

back in the rvc 1 days it took 8 hours to train a decent model

[indigo anchor]

which doesn't sound that great

[pulsar flume]

Ouch

[indigo anchor]

you can load models into it but it insists on using 0-shot tts

[pulsar flume]

I guess people then also didnt know much about the training statistics so they would undertrain or overtrain

[pulsar flume]

I mean it seems like the WebUI can do inferencing and training with models

I'll give it a try

[feral halo]

Ayo? @pulsar flume level 3 !!!

[timid musk]

now people just overtrain their models

because realistically

it’s really hard to make them sound worse with overtraining

[pulsar flume]

A model I had recently was 1000 epochs

[timid musk]

all your really doing is wasting time at a certain point

[pulsar flume]

Yeah it seems RVC when overtraining isnt introducing much noise now

[versed jetty]

what

[pulsar flume]

When I first used RVC (abt a year or 2 ago) overtraining killed models

[versed jetty]

,.,.,,..,.,.,.,.,.,.,.,.,.,.,.,.,.,.,,.,.,.,.,.,

[timid musk]

rvc didn’t exist 2 years ago

[versed jetty]

aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaabbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbcccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccdddddddddddddddddddddddddddddddddddddddddddeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeefffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffgggggggggggggggggggggggggggggggggggggggggggggggggggggggggggghijk

[pulsar flume]

Must of been a year ago then I have terrible memory lol

[timid musk]

well a year ago people were using SVC

[pulsar flume]

Tf you having an aneurism abt

[timid musk]

which was b a d

[pulsar flume]

That rings a bell

[versed jetty]

what is rvc

[timid musk]

retrieval voice-based conversion

[indigo anchor]

is there a way to determine what would be overtraining? my first thought was a ratio between how much data you put in to its epochs. if that ratio goes lower than [x] then it'd be overtraiing?

[twin fractal]

ai cover

[pulsar flume]

Hold on lemme fetch you a guide I saw. The images are dead so i'll send you the archive.org version

[indigo anchor]

it's useful for more than that

[twin fractal]

egril voice changer

[pulsar flume]

https://web.archive.org/web/20231013045136/https://rentry.org/RVC_making-models

Images are there in that snapshot

Goes into detail about overtraining

[indigo anchor]

the issue with doing that. if anyone knows the character of the voice you're trying to replicate, you've been busted

[twin fractal]

Paid models

[pulsar flume]

TL;DR

if G and D are going down that’s good
if they are mirroring each other, that’s good
if they stop going down check Mel and KL and if those are still going down good
basically if any of Mel, KL, G, D graphs are going down thats good.

[versed jetty]

iam the cookie monster

[indigo anchor]

what does that have to do with anything?

[frigid condor]

it means u commison someone to voice a model

[twin fractal]

No one will know what voice your replicatinh

[twin fractal]

ya

[indigo anchor]

what if that person gets commissions often?

[frigid condor]

than theyll be really popular ig

[twin fractal]

https://tenor.com/view/america-ya-gif-15374592095658975433

[twin fractal]

money 🤑

[versed jetty]

https://tenor.com/view/cookie-cookies-gif-19170472

[timid musk]

yes but here at ai hub

checks legal notes

[versed jetty]

iam the coookie monster

[timid musk]

we do not condone or facilitate creation of models created with copyrighted content

[twin fractal]

🤓

[versed jetty]

https://tenor.com/view/sesame-street-cookie-monster-cookies-hungry-gif-3504927

[frigid condor]

so vocaloid and other vocal synths are free game

[versed jetty]

iam the cookie monster

[frigid condor]

?

[indigo anchor]

no. their EULA's will likely prevent you from doing that too

although it's different depending on the company, so keep a note

[frigid condor]

ive been making tons of vocaloid models

[timid musk]

that’s on the user who uploads the model, not us

we are a public forum

😎

[frigid condor]

crypton ninjas gon get me

[pulsar flume]

But anything that isnt shared is uh.... not known about

[indigo anchor]

ok that's a fair point

[frigid condor]

but all my models are for personal use so

[timid musk]

personal use models will never and can never be restricted

[pulsar flume]

I love davinci resolves fusion.

[frigid condor]

with that said arent vocaloids sposed to be like yknow.. free ffor all ,anyone use their voice and character in any works and shit

[versed jetty]

https://tenor.com/view/cookie-monster-eating-cookie-hungry-yum-gif-9867685

[frigid condor]

they HAVE made it clear we cant make custom UTAU voicebanks of them like Miku

honestly idk where we draw the line

[versed jetty]

iam the cokie monster

[errant sphinx]

BUCKSHOT

[versed jetty]

cookie monster is the best in cookie eating

[indigo anchor]

most of my experience with vsynths (getting into synthv a year ago), most of them have a clause in their eula that says they can't be used to train singing/speech synthesis models

and also your statement on them being free for all is not entirely true. some models like fukase required permission from the developers for commercial use cases

[frigid condor]

that is... incredibly complex

[pulsar flume]

Can I ask what tf is a vocaloid haha

[indigo anchor]

to make things easier, "commerical use" was limited to things like posting music made with it on distribution sites like spotify or apple music. posting it to youtube wouldn't count as commercial use

[storm meadow]

@quartz roost

[frigid condor]

vocal synthesiser sofware

[storm meadow]

do your thing

[pulsar flume]

Ohhhhh voice manipulation?

[frigid condor]

i mean ,thats one way to put it

[pulsar flume]

I think I know whatchu mean lol

[sinful horizon]

Wenas Wenas

[indigo anchor]

it's a vsynth software made by yamaha, the same guys who make instruments, motorcycles, and hardware synthesizers. it's essentially their crack at providing vocals to people who don't know how to sing. on top of that, everyone who's on it agreed to be on there which is a plus

[coarse vessel]

vocal synthesizer, similar to TTS but with editable notes

[pulsar flume]

I see!

[frigid condor]

TTS but they can sing

[indigo anchor]

this is where hatsune miku originated back in 2007

[pulsar flume]

Yamaha make absolutely everything

[pulsar flume]

So the stuff websites like SUNO use to get the vocals?

[indigo anchor]

no

[pulsar flume]

Im confused sorry haha

[indigo anchor]

they contact the voice providers and they'd fly them out to their studio to record

[frigid condor]

a Voice Provider records voices which are than used to create Voicebanks, these are than used by creators around the world to make music

its like RVC training but not really

[indigo anchor]

for AI banks at least

[feral halo]

Ayo? @indigo anchor level 5 !!!

[indigo anchor]

for synths like traditional vocaloid or utau, they record the sounds individually and string them together through concatenative synthesis

[coarse vessel]

from traditional voicebanks to as AI models

[indigo anchor]

like sentence mixing in youtube poops

[pulsar flume]

Ohhhh so its basically a collection of voices that people can use and then the vocal synth software can manipulate that to make different sounds n shit?

Please say I got it semi right xD

[whole patrol]

и так

[frigid condor]

as far as i know, Yamaha only gives the Voicebank creattion tool to only specific brands, unlike UTAU whcih anyone can make their own Voicebanks with their own voice

[frigid condor]

yes

[quartz roost]

theres tools to do this automatically

[pulsar flume]

WhsiperX didnt do it right

I tried that first

[frigid condor]

you type あ and choose which note that あ is going to be sung in, u do this for every vowels/consoants to create comprehensible words

like a normal music intstrument but with words

[indigo anchor]

ai banks are a little different. you record the provider singing, then you determine where in the audio is what phoneme being sung (this process can take hours), and then you train them like most other ai models

[pulsar flume]

Ohhh so your basically phonemizing it

[frigid condor]

basically

[pulsar flume]

I think I get it now

[indigo anchor]

from my limited experience with diffsinger (utau but for ai banks), you'd need like an hour of data for english to get decent results

[pulsar flume]

Well I have never seen this import error from numpy

[quartz roost]

im using whisper v3 in "whispering tiger" with a few adjustments and its pretty close

[quartz roost]

Voice synthesizers specifically made for singing. There's thousands of them. Both free and paid.

[pulsar flume]

I mean it got most of it but the timecoding was out of wack and it was all bunched together so it was a huge block of subtitles

[quartz roost]

https://github.com/Sharrnah/whispering-ui

i use this, most languages

[pulsar flume]

Oh damn that actually looks decent lemme see if it works

[feral halo]

Ayo? @pulsar flume level 4 !!!

[quartz roost]

export as .csv

[quartz roost]

i use it as a realtime translator

[pulsar flume]

See thats handy. So you can pipe in audio from VB Cable and it'll translate?

Say in a VC

[quartz roost]

connected it to a real phone number and used it while traveling in a foreign country

it kinda worked

[pulsar flume]

Using VoIP?

[quartz roost]

i even loaded my favorite voice into it

[quartz roost]

yeah, a google voice number because it was international

[pulsar flume]

Oh thats cool asf

[quartz roost]

the new samsung galaxy advertised a similar feature. but it performs about the same while this is free open source

[pulsar flume]

I mean samsungs stuff is cool n all but yeah pretty much everything they have, theres a OSS version

[languid musk]

.

[versed jetty]

yes very cool man

أجل انه رائع

[quartz roost]

looking into it more i saw they are providing it for free for the new phone buyers but it looks like they WILL CHARGE MONEY for it later

i cant blame them though, running this realtime uses a LOT of processing power

[pulsar flume]

Yeah same as Microsofts Copilot which died in like 3 months

[quartz roost]

uses 6gb+ vram

[versed jetty]

مرحبا welcome

[pulsar flume]

Thats not terrible

[quartz roost]

V3 large 16 bit works decent

[pulsar flume]

I have a 3060 12gb model will it work decently enough with V3?

[quartz roost]

yeah, that will run realtime. i prefer using DEEPL Api