#✨│ai-help

1 messages · Page 179 of 1

knotty moth
#

do you mean the old one that doesn't even have rmvpe?

#

-gui

azure marshBOT
willow sage
#

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

ionic crag
rare gobletBOT
#

Ayo? @ionic crag level 1 !!! lfg

nova geode
#

Hey guys I need help with RVC I downloaded it and extracted the files to the correct locations on it but it keeps saying error when I try to convert

#

I’m using rvc WebUI

odd shale
#

If it's the second case, you can just use Applio Colab.

nova geode
#

Use models

nova geode
odd shale
#

There you go

nova geode
#

Ah

#

how do i install it?

odd shale
#

All you have to do is run certain cells and that's it.

nova geode
#

its still a ai voice changer right?

rare gobletBOT
#

Ayo? @nova geode level 3 !!! lfg

odd shale
nova geode
#

yeah

#

i think

odd shale
#

Because W-Okada does that.

#

Not RVC.

nova geode
#

so whats using models in real time?

odd shale
nova geode
#

yeah thats it

odd shale
#

-rt

azure marshBOT
# odd shale -rt

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

odd shale
#

Read Deiteris' guide.

nova geode
#

ah ok thank you

odd shale
#

I personally recommend Deiteris fork.

nova geode
#

i can add voices to this right

raven ginkgo
#

-colabv

#

-colab

azure marshBOT
# raven ginkgo -colab
☁️ Google Colabs
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
ancient portal
#

can anyone send?

tight stag
#

is there a way to combine audio to audio and tts to audio? Some song covers are hard to do as they have overlapping vocals and would like to manually edit the text myself to correct somethings the ai is mishearing

rare gobletBOT
#

Ayo? @tight stag level 2 !!! lfg

tight stag
#

my understanding is, rvc use voice to text then text to speech again correct more or less

#

so basically manually fix the text after voice to text

brittle wing
#

best way to currently train a model with raw vocals for music? had a collab but it doesnt work anymore

odd shale
#

-kaggle

azure marshBOT
tranquil blaze
#

Is it still improving or is it mode collapse?

odd shale
tranquil blaze
odd shale
#

Altho you could test the pth's that match these two peaks.

brisk quarry
#

No

odd shale
#

How'r you?

brisk quarry
#

No.

#

Im ugl

#

Dont look at me

odd shale
tranquil blaze
#

And speaking of this, when I have to test pth's I have to pick the lowest smoothed value, and not the lowest actual value, right?

brisk quarry
odd shale
tranquil blaze
rare gobletBOT
#

Ayo? @tranquil blaze level 5 !!! lfg

tranquil blaze
brisk quarry
tranquil blaze
brisk quarry
#

We all knew he should have made a movie

#

Shouldnt*

tranquil blaze
odd shale
# brisk quarry

Nep, if you wanna share memes you can do it on the verified channel, this channel is for helping.

brisk quarry
#

No...

#

That goofy ahhh springlock

odd shale
#

And how you did clean it.

tranquil blaze
# brisk quarry What about fnaf tho

Yeah, that was good, but it fit the style of the game imho, minecraft is too cartoon-y for it, they should've made something like story mode but as a film or something like that

brisk quarry
#

I miss story mode

tranquil blaze
odd shale
tranquil blaze
# brisk quarry I miss story mode

I actually never played it, but watched a couple of episodes of a playthrough when I was little, I'll catch up on it though because it's good :)

brisk quarry
#

@tranquil blaze

maxresdefault_8.jpg
tranquil blaze
#

I actually found out about it when I searched "minecraft" on steam (because I saw that my cousin bought every game from there, so I thought it was on there too lol) and that came up hahaha :P

brisk quarry
#

However, both seasons of the game became unplayable due to the closure of Telltale Games in late 2018, causing the game to be ultimately discontinued on June 25, 2019.

tranquil blaze
brisk quarry
#

You dont actually

#

How are you gonna play a game that do not exist

tranquil blaze
brisk quarry
#

The hardware optimization will not exist

tranquil blaze
odd shale
#

This channel is meant for helping about RVC.

brisk quarry
tranquil blaze
tranquil blaze
gusty loom
#

Guys, Im trying to use a ai model I found from one of the channels here but when I paste the link provided into the box it comes with an error, am I doing it wrong?

brisk quarry
#

No

mystic shell
#

Whenever I record something the first 2-3 seconds have random distorted robotic sounds but After that it works and sounds good. Even when I dont speak directly it makes distortion sounds before it works

cinder stag
#

applio was just working fine and now all of a sudden, all of my audio output comes out slowed down/pitched down. but i haven't messed with the actual pitch. i've changed models and songs, same result.
idk if this makes sense? but please lmk if anyone has had the same issue and has a fix

forest pewter
#

what UVR process method is the most consistent?

cinder wyvern
#

one of my voice models took like 5 minutes to download into the google colab gradio, and when i try to infer with it i get this error

#

Traceback (most recent call last):
File "/usr/local/lib/python3.10/dist-packages/gradio/queueing.py", line 532, in process_events
response = await route_utils.call_process_api(
File "/usr/local/lib/python3.10/dist-packages/gradio/route_utils.py", line 276, in call_process_api
output = await app.get_blocks().process_api(
File "/usr/local/lib/python3.10/dist-packages/gradio/blocks.py", line 1923, in process_api
result = await self.call_function(
File "/usr/local/lib/python3.10/dist-packages/gradio/blocks.py", line 1509, in call_function
prediction = await anyio.to_thread.run_sync(
File "/usr/local/lib/python3.10/dist-packages/anyio/to_thread.py", line 56, in run_sync
return await get_async_backend().run_sync_in_worker_thread(
File "/usr/local/lib/python3.10/dist-packages/anyio/_backends/_asyncio.py", line 2177, in run_sync_in_worker_thread
return await future
File "/usr/local/lib/python3.10/dist-packages/anyio/_backends/_asyncio.py", line 859, in run
result = context.run(func, *args)
File "/usr/local/lib/python3.10/dist-packages/gradio/utils.py", line 832, in wrapper
response = f(*args, **kwargs)
File "/content/program_ml/program_ml/core.py", line 119, in run_infer_script
infer_pipeline.convert_audio(
File "/content/program_ml/program_ml/rvc/infer/infer.py", line 317, in convert_audio
self.get_vc(model_path, sid)
File "/content/program_ml/program_ml/rvc/infer/infer.py", line 907, in get_vc
self.setup_network()
File "/content/program_ml/program_ml/rvc/infer/infer.py", line 943, in setup_network
self.tgt_sr = self.cpt["config"][-1]
KeyError: 'config'

tranquil blaze
#

I trained for 500 epochs, should I use that checkpoint or something from earlier than that point?

#

Or even this point here, before the first time the curve went up

tough lark
#

does anyone know how to find social media by photo?

knotty moth
tranquil blaze
hybrid rune
#

what do I do from here? @royal jacinth

image.png
sonic agate
tawdry shell
#

what does it mean when the error says: "index -1 is out of bounds for axis 0 with size 0"???

urban turret
#

I tried a backup of Applio 3.0.8 I’ve had, and it worked normally, but I’d still like to use the updated version.

tawdry shell
#

Help, I don't know what this error means haha

image.png
#

Is it something with the model? I've never encountered this issue before with other models

tawdry shell
#

why does it say this when I'm trying to download applio? T^T

image.png
knotty moth
#

sorry forgot to say but the latest Applio has feature to apply reverb, not sure if the colab has that too

stone urchin
#

hello!

#

I've been trying to use the google collab for a while but it seems to be broken , I wanted to know if there is an alternative to use the application .

knotty moth
stone urchin
low shard
#

do you wanna inference (use models) or train (make models)?

stone urchin
rare gobletBOT
#

Ayo? @stone urchin level 1 !!! lfg

stone urchin
#

but if you have a guide for making models , may it can be useful

low shard
# stone urchin use models

use ilaria rvc zero that is a zerogpu huggingface space (A100 gpu faster than colab's T4 gpu, so its faster than doing it on colab)

brave garnetBOT
low shard
stone urchin
#

@low shard thank you! and Sul-Sul!

low shard
knotty moth
rare gobletBOT
#

Ayo? @knotty moth level 38 !!! lfg

stone urchin
low shard
stone urchin
#

Ilaria give me this error xd

image.png
knotty moth
low shard
#

tbh that seems like the No GPU Avaible for 60s or either a too long file,
unless you already used it a some times

Google Docs
Ilaria RVC Zero

Table Of Contents Introduction (with website link) Model Loader (Download & Upload) Inference (use RVC AI Voice Models) Ilaria TTS Settings (Inference) Troubleshooting “No gpu is available for you for 60s” Introduction (with website link) Ilaria RVC Zero, is an RVC (Retrieval-based Voice Co...

stone urchin
#

it works now

stone urchin
#

im using blinding lights so its 2:55 long

proper shale
low shard
low shard
# stone urchin which it's the limit ?

can't say exactly but zerogpu inference limit is 300 seconds (in terms of time the A100 gpu takes to process your audio, not your audio lenght), but i think like 30 mins

stone urchin
#

@proper shale is it true that he composed the soundtrack for sonic 3 ?

stone urchin
low shard
proper shale
# stone urchin <@622856015444049937> is it true that he composed the soundtrack for sonic 3 ?

Actually, no.

The thing is, they were apparently planning to hire Michael to do some songs, but it failed because of the 1993 accusations and all that. The person who made the supposed "MJ" tracks was actually Brad Buxer, which was one of Michael's producers/studio engineers and was a keyboardist for his tours. Brad is responsible for some of MJ's sound after 1991, and you can kind of hear MJ's sound because it's Brad who made them.

stone urchin
#

now i got error on the vocal separator

image.png
#

i wanna to clean some audio but every model give error

knotty moth
low shard
low shard
raven ginkgo
#

any tips on how to make a voice model sound less robotic and breathy?

vapid niche
#

guys can you help i have 4060 laptop and i core i5 can you help please

tame widget
#

guys i can't find a good rvc trainer

#

i have a good laptop

low shard
low shard
tame widget
#

7300

rare gobletBOT
#

Ayo? @tame widget level 1 !!! lfg

obtuse galleon
#

anyone knows how to solve this? its the UVR_UI

image.png
mortal robin
#

I hope you're doing well. Sorry to bother you, but I’m reaching out because I need one of you to help me create a voice model pls

tender sand
#

ive got a voice model downloaded, but im not sure how im supposed to import it and be able to change my voice as the guides i looked at did not specify it

odd shale
odd shale
odd shale
tender sand
odd shale
#

I would highly suggest you to use Deiteris' w-okada fork.

#

-rt

azure marshBOT
# odd shale -rt

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

tender sand
#

i downloaded it from okada

odd shale
#

W-Okada is for using models on realtime.

odd shale
obtuse galleon
tender sand
rare gobletBOT
#

Ayo? @tender sand level 1 !!! lfg

odd shale
#

Eddy ven a ayudarlo.

viscid moss
#

Use Hugging Face or UVR5 NO UI instead

obtuse galleon
rare gobletBOT
#

Ayo? @obtuse galleon level 2 !!! lfg

viscid moss
simple mortar
#

Hello I have an rtx-4070 and I want to make some models. However I have no idea where to start. I'm Python literate and I have WSL and docker so if needed I can run on those for GPU stuff.
quest #1 how do i train one?

odd shale
#

-rvc

azure marshBOT
simple mortar
#

ty

#

man I really should have asked here first lol. I was trying to code my own with torch in python haha

#

after training it i could not get it to load in the voice changer haha

#

wow this is so easy. 500 epoch should do it. Yo Leo dont know if you were being frank or jerkish ether way thank you. I'm off to the races !!! 🙂

rare gobletBOT
#

Ayo? @simple mortar level 1 !!! lfg

simple mortar
#

Is there a way to view the output? it says processing but no terminal.

odd shale
odd shale
#

You can do that after fully finishing the training of a model and testing it with an audio

simple mortar
#

no i just mean the terminal. like in another model i made for game audio detection it would show the training details like accuracy, val_accuracy, loss, val_loss.

#

do you know if over fitting is an issue when making voice models?

#

nevermind i found it in the docs

#

TensorBoard
-Use the model that was closest to before the OT point.
For more information, check the guide for your RVC in this website.

how do i get the model before OT are they stored some place?

polar plaza
#

Why is it doing this

#

It was working fine like 5 minutes ago

#

I’ve refreshed it and everything

low shard
# polar plaza

Did you get like an “No gpu is available for you for 60s”

simple mortar
#

is it over fitting?

image.png
rare gobletBOT
#

Ayo? @simple mortar level 2 !!! lfg

odd shale
simple mortar
#

can i not stop it early?

#

okay found them it saves ever 5 lol and i found the output folder

simple mortar
# 
RVC:: slotInfo.modelFile G_9345.pth
[Voice Changer] post_load_model ex: 'config'
Traceback (most recent call last):
  File "restapi\MMVC_Rest_Fileuploader.py", line 92, in post_load_model
  File "voice_changer\VoiceChangerManager.py", line 170, in loadModel
  File "voice_changer\RVC\RVCModelSlotGenerator.py", line 37, in loadModel
  File "voice_changer\RVC\RVCModelSlotGenerator.py", line 43, in _setInfoByPytorch
KeyError: 'config'```
brittle wing
#

Can someone help me transcribe a song?Most of the lyrics are clear to me, it's a rare SoundCloud song by an YouTuber rapper, I can send the file and the transcribed lyrics.I'm planning on making a lyric video, so I need someone's help.I need a native English speaker, thanks.
Also to correct some of the parts that are wrong etc.Also don't tell me to pay an AI or stuff, I have already tried to trasncribe with ai, helped to some extent but human ear is something else

brittle wing
#

How can i upload a model to weighs

hollow wasp
#

how do I use the pretrained model with the D and G thing

steep cipher
#

what is the best realtime voice changer ?

brittle wing
#

An error occurred extracting the index: need at least one array to concatenate
If you are running this code in a virtual environment, make sure you have enough GPU available to generate the Index file.

#

HELP

#

applio rvc google colab btw

brittle wing
knotty moth
brittle wing
#

Mhm I can dm you some of my ai covers (not that I won't redo them) and tell me if they're good or not

knotty moth
brittle wing
knotty moth
hollow wasp
#

@knotty moth could you help me with the G and D file of the pretrained model? I uploaded them on rcv web ui and did some training with it, but idk where to get the weight

brittle wing
knotty moth
brittle wing
#

Also Uvr architecture is always TOP IMO.

brittle wing
#

-colab

azure marshBOT
# brittle wing -colab
☁️ Google Colabs
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
rare gobletBOT
#

Ayo? @brittle wing level 4 !!! lfg

loud nacelle
#

what is the alt for using the voice changer without GPU

#

colab don work anymore

knotty moth
# loud nacelle what is the alt for using the voice changer without GPU

even modern laptops can actually run the fork voice changer locally in cpu mode: https://rentry.co/forkvoicechangerguide

Guide for deiteris' optimized W-Okada RealTime Voice Changer Client...

Guide style is in the same as Blanc_dot's. Thanks Blanc_dot for corrections. Most technical information comes from deiteris.
Last update August 30th, 2024: New b2309 version
Translations added for:
German: https://rentry.co/ForkVoiceChangerGuide_de
Turkish: https://rentry.co/ForkVoiceChangerGuide...

slate zodiac
#

Is it normal for UVR to convert my 48kHz mono wav into a 44.1kHz stereo wav? I'd like to keep the original sample rate. My source is definitely 48, verified with a spectrogram

rare gobletBOT
#

Ayo? @slate zodiac level 1 !!! lfg

brittle wing
#

-guide

azure marshBOT
brittle wing
#

How much hop length for rvmpe?

#

128:2=64?

rare gobletBOT
#

Ayo? @frail kestrel level 1 !!! lfg

knotty moth
#

however, you can try the slowdown trick by interpolate sample rate of the input file to 44.1k or lower, and then revert the output sample rate back
although that may possibly cause different result and may not work on deep male voices

glacial pollen
# knotty moth however, you can try the slowdown trick by interpolate sample rate of the input ...

Wouldn't advice it really, interpolation of 44.1 to 48 is worse than having plain 44.1 tbf, that's just simply due to the fact no interpolation or math in approaches like that is error-prone free

Comparing 44.1 and 48 has virtually no audible difference unless you're sensitive to high-ends and high freq brightness differences ( but again, if one doesn't work on audio daily and has sufficient gear, it's something an user on avg won't notice

simple mortar
#

yesterday it was working today i'm just getting 1 dot???

image.png
simple mortar
#

i see every 200 steps.

image.png
odd shale
odd shale
knotty moth
balmy citrus
#

Can i do something if the model cuts some words ? maybe a setting problem ?

obtuse stream
#

hi anyone help me what is the difference between the two? 64bit float and pcm 32?

Ultimate_Vocal_Remover_2024._09._12._15_32_55.png Ultimate_Vocal_Remover_2024._09._12._15_33_06.png
odd shale
#

This tends to happen if you use audio samples with overlapping vocals/stuff

slate zodiac
knotty moth
knotty moth
glacial pollen
lavish escarp
#

-colab

azure marshBOT
# lavish escarp -colab
☁️ Google Colabs
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
calm locust
#

cannot use rvc, but can use beatrice . Why?

odd shale
pastel oak
odd shale
brittle wing
#

i need help.. aicovergen in huggingface gives error and doesn’t work anymore :’I what other similar generators can i use?

brittle wing
azure marshBOT
brittle wing
#

i recommend using ilaria rvc zero

odd shale
brisk quarry
#

I reccommend not using rvc at all

azure marshBOT
odd shale
knotty moth
#

be specific, which space link?

open musk
#

does anyone know site where you can put .index and .pth files i didn't do that in a long time and i forgot

odd shale
low shard
fading summit
#

does anyone know where i can get the audio sample that says "my model was made with whatever the name of the model is"

low shard
#

forks like Applio & Ilaria RVC have built in tts

fading summit
#

ive seen alot of people us the same audio sample tho

#

word for word

latent pumice
#

Question, the VoiceChangerType, How do i know which is better? Or should i stick to RVC?

leaden wedge
#

hey gang. Is there a CLI utility to blend two voice models?
I'm using the Mangio RVC Fork on a P4000 machine on Paperspace. Using the CLI tool I've been able to train my own models and run inference, but I don't see an option to do the "ckpt Processing" that you can do thru the Web UI

odd shale
#

-rt

azure marshBOT
# odd shale -rt

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

obtuse ore
#

Hello, does this only support NVIDIA gpu? I have a an AMD Raedon RX 7900 XTX and it is very choppy below 100 chunks. It isn't choppy at all at around 192 chunks but very delayed.

brittle wing
#

How much crepe hop length for rvmpe?

analog obsidian
#

the setting only applies if you use mangio, if not, it doesn't do anything

brittle wing
#

128:2=64?

analog obsidian
brittle wing
analog obsidian
#

but if ur too worried, set it to 64

brittle wing
analog obsidian
#

rmvpe hop length is hardcoded, so it cannot be changed

analog obsidian
slow creek
#

Hi! New to using RVC. I've been using Bark and models in a npz file to generate ai voices.

I tried using TTS Generation WebUI locally, but I keep running into issues. So, now I'm using Bark in a Jupyter Notebook. I could generate my own models, but they take forever. The file type of my models are npz.

Is there a way to convert RVCs to npz files, or use Bark with RVCs?

proper shale
#

You could probably use generated audios from Bark and convert the voice with RVC, though

odd shale
proper shale
odd shale
brittle wing
analog obsidian
brittle wing
analog obsidian
brittle wing
analog obsidian
#

for mangio training a hop of 64 or 32 is fine
anything below those numbers also works but there is barely to 0 audible difference

brittle wing
cosmic moon
#

when i have a rvc upload and the base ones dont work how do i ifx

slow creek
elder halo
#

is rx 6700xt 12 gb good for rvc?

proper shale
proper shale
grave river
#

I have a 6600AMD why is the voice changer using 90 to 100 percent of my cpu?

rare gobletBOT
#

Ayo? @grave river level 2 !!! lfg

slow creek
rare gobletBOT
#

Ayo? @slow creek level 1 !!! lfg

proper shale
# grave river I have a 6600AMD why is the voice changer using 90 to 100 percent of my cpu?

You're probably using the wrong version of Okada for your GPU.

Here's the guide for the fork that has better performance for AMD cards: https://rentry.co/forkvoicechangerguide

Guide for deiteris' optimized W-Okada RealTime Voice Changer Client...

Guide style is in the same as Blanc_dot's. Thanks Blanc_dot for corrections. Most technical information comes from deiteris.
Last update August 30th, 2024: New b2309 version
Translations added for:
German: https://rentry.co/ForkVoiceChangerGuide_de
Turkish: https://rentry.co/ForkVoiceChangerGuide...

proper shale
elder halo
proper shale
#

The 3050 is garbage and the 6700 is an AMD card so none of them are ideal for RVC

#

Aim for a 3060, 12GB, should be enough for realtime, training models, all that stuff

elder halo
lavish eagle
#

i forgor where to get the ai lmao

knotty moth
brittle wing
grave river
#

Thanks now it’s not abusing my cpu

#

👍😊

knotty moth
elder halo
rare gobletBOT
#

Ayo? @elder halo level 1 !!! lfg

knotty moth
elder halo
brittle wing
#

-hf

azure marshBOT
edgy monolith
#

tried continuing a model that i was training earlier

#

am i able to fix this?

knotty moth
edgy monolith
knotty moth
potent helm
#

What pretrain would be the best for raw studio vocals around 10 minutes that cutoff at 48khz

edgy monolith
knotty moth
edgy monolith
#

uh oh

#

i had mine at 8

knotty moth
# edgy monolith i had mine at 8

it has reportedly crashed on feature extraction with crepe at any hop length and all cpu threads used, regardless if it's intel or AMD cpu
but that should be fine if you have done so

brittle wing
#

I'm not sure why there's no AI model out of 90's Snufkin. I tried asking @lunar needle about him, but he still won't listen to me at all.

low shard
#

You can search rvc ai voice model at:

if there isnt one, you can:

untold kernel
#

Hi guys 😊. Quick question

If i create a voicemodel in lets say the scale of F#. Will it pitch well in other keys or is it preferred to export it in multiple scales to fit the input audio better? I made a lot of VM’s in the same key (since it is easier) but got in doubt if it is a good idea 😊

quiet surge
#

how do i use the d and g files toegther?

quiet surge
knotty moth
quiet surge
knotty moth
quiet surge
rare gobletBOT
#

Ayo? @quiet surge level 1 !!! lfg

knotty moth
low shard
#

i was just gonna send that lol

quiet surge
low shard
quiet surge
#

this one

knotty moth
quiet surge
#

oh mb guys

low shard
untold kernel
rare gobletBOT
#

Ayo? @untold kernel level 1 !!! lfg

knotty moth
untold kernel
fair ravine
#

hey guys imma using uvr 5 but I can't select MDX23C (it doesn't show it) but it shows (uvr-mdx-net inst hq 3). What can I do?

knotty moth
fair ravine
#

@knotty moth where is "process method" option on this program?

rare gobletBOT
#

Ayo? @fair ravine level 2 !!! lfg

fair ravine
#

resolved

wraith coral
#

Traceback (most recent call last):
File "C:\Users\kkkkkkk\Documents\RVC1006Nvidia\gui_v1.py", line 874, in <module>
gui = GUI()
File "C:\Users\kkkkkkk\Documents\RVC1006Nvidia\gui_v1.py", line 114, in init
self.launcher()
File "C:\Users\kkkkkkk\Documents\RVC1006Nvidia\gui_v1.py", line 400, in launcher
self.event_handler()
File "C:\Users\kkkkkkk\Documents\RVC1006Nvidia\gui_v1.py", line 431, in event_handler
self.start_vc()
File "C:\Users\kkkkkkk\Documents\RVC1006Nvidia\gui_v1.py", line 558, in start_vc
self.gui_config.samplerate = self.rvc.tgt_sr
AttributeError: 'RVC' object has no attribute 'tgt_sr'

odd shale
patent trellisBOT
# odd shale !howtoask

How To Troubleshoot AIHC_WaitWhat

__**GIVE CONTEXT.**__ 📝
  • Don't simply mention your issue, like "my rvc is not working".
  • Describe the step you are on, what you're trying to do, the RVC you're using, a screenshot, etc.
  • The more context, the better.
__**BE POLITE.**__ <:matsuripray:1159685390156967936>
  • Don't be desperate. You can ping a Helper, but if they ignore, they aren't available/don't know the answer.
  • It's okay if you're frustrated, but don't take it into this server.
  • Don't DM without prior consent.
__**BE PRODUCTIVE.**__ 🤝
  • Don't ask for every little instruction. Put your own effort & test things by yourself.
  • Don't ask to ask.
  • Check if your answer is a Google search away/on our guides website.
If you've changed your mind about your message, feel free to delete/edit it.
lavish escarp
#

-colab

azure marshBOT
# lavish escarp -colab
☁️ Google Colabs
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
amber charm
#

hey ik it's probably been asked before, but I'm training a 48K Mangio Crepe model, is there any pretrain that can help the process?

proper shale
#

just use original

#

but tbh, this doesn't seem like a good idea in general, unless you have a suuuuper clean dataset the advantages will basically be minimal, and the disadvantages will be many

32k and rmvpe is just fine for most datasets, and may not have the same problems as a 48k crepe model

red kayak
#

@analog obsidian stop typing an essay

analog obsidian
#

i was about to say that mangio is less accurate than rmvpe xD

red kayak
analog obsidian
red kayak
raven ginkgo
#

-colab

azure marshBOT
# raven ginkgo -colab
☁️ Google Colabs
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
raven ginkgo
#

-kaggle

azure marshBOT
raven ginkgo
#

-local

azure marshBOT
raven ginkgo
#

-audio

azure marshBOT
carmine hearth
golden walrus
#

guys. i tried to use applio to train but this pop up

image.png
#

and this is the cmd look like

image.png
#

😦 is there any guide to fix it ?

analog obsidian
#

Make sure the datasets lacks pops in those cut segments

#

Pops sounds are like crackling noise

#

Like when ur lagging in a game

#

Usually loud

#

You can fix the crackling sound in rx10 using mouth de clicker

#

Its fine to have cut sentences in the dataset, rvc already does that when finds long sentences

#

whats bad is the crackling/pops u get when you cut the sentence manually

tropic kelp
#

Why does voice models have crdf virus in them

analog obsidian
#

If you don’t fix the pops/crackling then rvc is going to occasionally add the crackling noise to the model

rare gobletBOT
#

Ayo? @brittle wing level 8 !!! lfg

analog obsidian
#

Yea if the crackling only happened once it most probably ignored it

#

Have you tried uvr denoise and mel denoiser?

tropic kelp
analog obsidian
analog obsidian
#

it should be easy for you to identify the crackling, it is a loud sound that can be heard without any problem, in your case, the crackling should sound at the moment where you cut the audio.

#

this is how crackling looks in a spectogram

image.png
#

rx allows u to fix the crackling using the de-crackle module or mouth de clicker

image.png
#

you can use spek

#

if you have crackle it should look like this

image.png
#

the cut audio also may have clicking noise that is hard to spot in the spectogram, but you can also hear it

#

no prob, in case you don't how to spot clicking noise, it sounds like this

knotty moth
#

fyi crackles like shown in above examples are usually caused by clipping, so you can try de-clip and de-crackle (either only the latter or both) in izotope RX

analog obsidian
#

you can fix that in izotope rx with the module mouth de clicker or de-click
so first picture is the clicking, notice is a large yellow line, after mouth de clicking is applied the yellow line dissappears and you cannot longer hear the click

image.png image.png
analog obsidian
#

basically the pop u get after cutting any audio

Captura_de_pantalla_2024-09-14_115550.png Captura_de_pantalla_2024-09-14_115557.png
#

but yea de-crackle or de-clip fixes those

#

nails bro is writting an essay

knotty moth
analog obsidian
knotty moth
analog obsidian
knotty moth
elder halo
knotty moth
brittle wing
knotty moth
elder halo
knotty moth
low shard
#

if its no where in all the sites i sent then u have only that options

neon island
#

Slightly dumb question: is there any way i can get a voice model of the opposite gender to match my original octave? (for like singing and whatnot, cuz i kinda wanna do my own singing for funsies lol)

#

maybe i'm not describing it quite correctly

rare gobletBOT
#

Ayo? @neon island level 1 !!! lfg

vivid sedge
#

I have a question

#

How do I train a rvc model? With 10000 epochs or 200 epochs?

marsh grove
#

-rvc

azure marshBOT
marsh grove
#

-colab

azure marshBOT
# marsh grove -colab
☁️ Google Colabs
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
loud tulip
#

i just updated applio and it stopped working and throws this erorr ```
pydantic.errors.PydanticSchemaGenerationError: Unable to generate pydantic-core schema for <class 'starlette.requests.Request'>. Set arbitrary_types_allowed=True in the model_config to ignore this error or implement __get_pydantic_core_schema__ on your type to fully support it.

If you got this error by calling handler(<some type>) within __get_pydantic_core_schema__ then you likely need to call handler.generate_schema(<some type>) since we do not call __get_pydantic_core_schema__ on <some type> otherwise to avoid infinite recursion.

For further information visit https://errors.pydantic.dev/2.9/u/schema-for-unknown-type

astral jungle
#

hi

odd shale
#

Personally i don't suggest using other pretrains.

#

Just use OG one instead.

#

Sadly i don't think there's a way to fix that.

#

You can also use RVC Mainline on kaggle.

#

-kaggle

azure marshBOT
slim owl
#

How do I make AI Voice Model?

odd shale
slim owl
odd shale
#

Depending if you wanna make a model of your own voice or a fictional character/singer/streamer

odd shale
azure marshBOT
odd shale
#

The one that says "docs aihub"

hot lagoon
#

can RVC AI Cover Maker work on paperspace?

odd shale
flint solar
#

why am igetting this error

odd shale
flint solar
#

😭

odd shale
#

Lo siento

flint solar
young zenith
#

what happened to Ilaria RVC ?

low shard
#

if u are talking about google colab its ded

low shard
brave garnetBOT
small ridge
#

please help

#

what happened to the google colab rvc

#

it keeps saying modulenotfounderror

ocean quail
#

How do you get a hugging face download link?

lethal shale
#

hellppp

#

i dont hear myself in vrc

brittle wing
#

Best inference settings?

brittle wing
#

-guides

azure marshBOT
brittle wing
#

-colab

azure marshBOT
# brittle wing -colab
☁️ Google Colabs
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
fair prism
#

does the new applio version give you guys an error or

vast fractal
#

does anyone know how to fix this i pressed process data and it said this
2024-09-14 23:55:38 | INFO | main | "runtime\python.exe" infer/modules/train/preprocess.py ""C:\Users\raypl\Downloads\Perch Perkins\PerchPerkinsDATASET.flac"" 40000 16 "C:\Users\raypl\Downloads\RVC1006Nvidia/logs/perch-perkins" False 3.0
['infer/modules/train/preprocess.py', 'C:\Users\raypl\Downloads\Perch', 'Perkins\PerchPerkinsDATASET.flac', '40000', '16', 'C:\Users\raypl\Downloads\RVC1006Nvidia/logs/perch-perkins', 'False', '3.0']
Traceback (most recent call last):
File "C:\Users\raypl\Downloads\RVC1006Nvidia\infer\modules\train\preprocess.py", line 11, in <module>
sr = int(sys.argv[2])
ValueError: invalid literal for int() with base 10: 'Perkins\PerchPerkinsDATASET.flac'
2024-09-14 23:55:39 | INFO | main |

rare gobletBOT
#

Ayo? @vast fractal level 7 !!! lfg

carmine hearth
#

Perch Perkins -> PerchPerkins

vast fractal
# carmine hearth Perch Perkins -> PerchPerkins

it still says 2024-09-15 00:34:29 | INFO | main | "runtime\python.exe" infer/modules/train/preprocess.py ""C:\Users\raypl\Downloads\Perch Perkins\PerchPerkinsDATASET.flac"" 40000 16 "C:\Users\raypl\Downloads\RVC1006Nvidia/logs/perchperkins" False 3.0
['infer/modules/train/preprocess.py', 'C:\Users\raypl\Downloads\Perch', 'Perkins\PerchPerkinsDATASET.flac', '40000', '16', 'C:\Users\raypl\Downloads\RVC1006Nvidia/logs/perchperkins', 'False', '3.0']
Traceback (most recent call last):
File "C:\Users\raypl\Downloads\RVC1006Nvidia\infer\modules\train\preprocess.py", line 11, in <module>
sr = int(sys.argv[2])
ValueError: invalid literal for int() with base 10: 'Perkins\PerchPerkinsDATASET.flac'
2024-09-15 00:34:30 | INFO | main |

#

stop spamming people

carmine hearth
#

\Downloads\Perch Perkins -> \Downloads\PerchPerkins

vast fractal
vast fractal
# carmine hearth Perch Perkins -> PerchPerkins

now when i tried to feature extraction i got 2024-09-15 00:41:12 | INFO | main | "runtime\python.exe" infer/modules/train/extract_feature_print.py cuda:0 1 0 0 "C:\Users\raypl\Downloads\RVC1006Nvidia/logs/perchperkins" v2
['infer/modules/train/extract_feature_print.py', 'cuda:0', '1', '0', '0', 'C:\Users\raypl\Downloads\RVC1006Nvidia/logs/perchperkins', 'v2']
C:\Users\raypl\Downloads\RVC1006Nvidia/logs/perchperkins
load model(s) from assets/hubert/hubert_base.pt
move model to cuda
no-feature-todo
2024-09-15 00:41:16 | INFO | main | ['infer/modules/train/extract/extract_f0_print.py', 'C:\Users\raypl\Downloads\RVC1006Nvidia/logs/perchperkins', '16', 'rmvpe']
no-f0-todo
no-f0-todo
no-f0-todo
no-f0-todo
no-f0-todo
no-f0-todo
no-f0-todo
no-f0-todo
no-f0-todo
no-f0-todo
no-f0-todo
no-f0-todo
no-f0-todo
no-f0-todo
no-f0-todo
no-f0-todo
['infer/modules/train/extract_feature_print.py', 'cuda:0', '1', '0', '0', 'C:\Users\raypl\Downloads\RVC1006Nvidia/logs/perchperkins', 'v2']
C:\Users\raypl\Downloads\RVC1006Nvidia/logs/perchperkins
load model(s) from assets/hubert/hubert_base.pt
move model to cuda
no-feature-todo

carmine hearth
vast fractal
brisk quarry
#

@glad zealot we are doomed

fair prism
#

because pydantic v2 installed instead of v1, applio no longer works

#

can someone tell me how to install v1 instead

brisk quarry
#

How comes i use it?

tame mica
#

@knotty moth hey vro

brisk quarry
#

@tame mica hello cutie

tame mica
fair prism
#

hacks

brittle wing
#

-colab

azure marshBOT
# brittle wing -colab
☁️ Google Colabs
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
brittle wing
#

-huggingspace

#

-realtime

azure marshBOT
brittle wing
#

-space

#

-hugging

#

-colab

azure marshBOT
# brittle wing -colab
☁️ Google Colabs
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
pastel oak
spice niche
#

How do I make an rvc voice model

low shard
low shard
low shard
haughty thistle
#

Hello everyone, I have a problem. I followed the tutorial and launched the 'go-web' folder, but after that, it opens a console and then nothing happens. It asks me to press a key to continue, but nothing happens except that the console disappears. Could you please help me?

pastel oak
spice niche
#

I know everything other than cpu and gpu 😭 I built my pc like 3 years ago

low shard
tired cloak
#

it tells me how to turn on running with the system python no MoPress any key to continue on a black screen, what should I do?

native axle
#

i keep getting value nan on every epoch

#

i got a 1650 with up to date drivers

#

is there any solve for this?

brisk quarry
#

get more vram gpu

low shard
#

its better u do it on cloud

native axle
#

n how can i do that

low shard
# native axle n how can i do that

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)
  • Lightning.ai (Kinda hard, needs login, no issue with web uis or anything, but only free 15 credits monthly):
  • Mainline (UI, No guide as of right now)
  • Applio (UI, No guide as of right now)
#

I'd personally suggest kaggle

knotty moth
native axle
low shard
tired cloak
rare gobletBOT
#

Ayo? @tired cloak level 1 !!! lfg

brisk quarry
#

@native axleif you send dataset i can train for you

native axle
rare gobletBOT
#

Ayo? @native axle level 1 !!! lfg

native axle
#

so i can train more models in the future

brisk quarry
#

you just cant train on a 1650

#

you need atleast rtx

native axle
#

i can train on an rtx

brisk quarry
#

yes

native axle
#

im just temporarily on a 1650

brisk quarry
#

then switch

low shard
#

But would use only when u got ur rtx

#

until then u need to use cloud for training

native axle
karmic fjord
#

-colab

azure marshBOT
# karmic fjord -colab
☁️ Google Colabs
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
karmic fjord
#

what happened to this?

#

help

low shard
#

its an old deleted google colab

#

what are u looking for?

odd shale
#

What you wanted to do?

#

Train models or use models?

spice niche
rare gobletBOT
#

Ayo? @spice niche level 1 !!! lfg

low shard
#

As you got a good PC, you can use RVC locally, you can choose between:

  • Applio: A fork of RVC with some extra features like Applio TTS, same quality tho
  • Mainline: The original RVC
spice niche
#

and can i make voice models with this and is there any tutorials?

low shard
spice niche
low shard
karmic fjord
# low shard where did u find that?

I used it for more than a year (best so far, personally) and it disappeared few weeks ago. I was asking if the link changed or the author removed it. Would be awesome if I found it exactly, or very similar one. Thanks in advance

spice niche
#

yeah this is the part im stuck on

image.png
low shard
low shard
spice niche
#

so do i put just "0" or "GPU 0"

karmic fjord
low shard
low shard
#

what do u want to do

karmic fjord
#

train, infer yk create models and and use them to produce AI voices

low shard
#

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)
  • Lightning.ai (Kinda hard, needs login, no issue with web uis or anything, but only free 15 credits monthly):
  • Mainline (UI, No guide as of right now)
  • Applio (UI, No guide as of right now)

I gave you all options, but for Training i would personally suggest any Kaggles, while for inference i would suggest Ilaria RVC Zero

Google Docs
Ilaria RVC Zero

Table Of Contents Introduction (with website link) Model Loader (Download & Upload) Inference (use RVC AI Voice Models) Ilaria TTS Settings (Inference) Vocal Separator (UVR) Troubleshooting “No gpu is available for you for 60s” Introduction (with website link) Ilaria RVC Zero, is an RVC (Re...

#

Ofc u can inference also on other things like on some google colabs like applio, but ilaria rvc zero has better gpu for free (only for inference)

karmic fjord
brittle wing
#

Best settings for inferencing?

odd shale
near gale
low shard
#

You can search rvc ai voice model at:

if there isnt one, you can:

#

i think they just scrape models so u should be able to find it in those sites too

knotty moth
knotty moth
paper dock
#

what sample rate pretrain should i use if most of my data look like this?

Screenshot_35.png
#

32k or 40?

analog obsidian
woeful shore
#

whats the best model for realtime

#

i remember it used to be rvc

ashen flint
#

Question do you guys know any group that has a song requester

#

Or cam you do it on this one?

#

@woeful shore

paper dock
#

i believe

tame mica
#

no it's diff-svc

carmine hearth
haughty thistle
karmic fjord
knotty moth
low shard
#

all youtube vids are outdated

#

therews the written guide in the link i sent u tho

karmic fjord
#

Well, I sent a close technerd friend close of mine the link, he'll upload a YouTube tutorial soon

storm hatch
#

hey folks I dont really want this message to get buried but are there any good tts voice cloners other than applio playground? for free at least.

#

either using rvc or a audio file upload

knotty moth
low shard
# storm hatch hey folks I dont really want this message to get buried but are there any good t...

There are different Text To Speech (TTS) AIs:

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

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

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

With RVC Models:

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

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

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

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

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

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

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

low shard
storm hatch
rare gobletBOT
#

Ayo? @storm hatch level 2 !!! lfg

storm hatch
#

thanks

#

fishspeech is kinda good actually, ill try gpt so vits on some time

low shard
knotty moth
#

what model were you using? I still hear the instrument residuals from the beginning
how about try either of these in mvsep.com?

image.png image.png
grim bay
odd shale
#

Please don't post copyrighted acapellas on the server.

grim bay
#

I will delete it

#

So, is there any ways to fix that part?

odd shale
grim bay
#

Any different models to separate the vocals and the song?

odd shale
#

Also, you must remove the backing vocals, reverb and echo.

grim bay
odd shale
grim bay
odd shale
polar totem
#

hey my thing just opens and closes

#

the start http

#

@ me if anyone knows

low shard
#

@river coral

#

-rt

azure marshBOT
# low shard -rt

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

low shard
#

1st link

#

ask in help channels next time

river coral
#

K

#

tysm

river coral
warped onyx
#

-colab

azure marshBOT
# warped onyx -colab
☁️ Google Colabs
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
forest pewter
#

tips on recording own vocals for training?

#

and later processing like noise reduction, what models generally work best

hollow sable
#

How do i fix my the crazy echo and the ai sounding super weird like no matter the model

#

I am ok w hopping on vc and sharing screen if it makes it faster

unreal surge
#

-rt

azure marshBOT
# unreal surge -rt

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

viral heath
#

i need hugginface space for RVC tts since 15.ai and uberduck died

knotty moth
viral heath
#

using ai models for text to speech

#

like Silero TTS telegram bot

knotty moth
viral heath
#

im new to this stuff

#

so, are there any spaces that can do it?

#

I need tf2 tts

knotty moth
subtle cedar
#

-applio

azure marshBOT
subtle cedar
#

-help

azure marshBOT
viral heath
#

I uploaded audio and skipped samples. no audio on prompt

#

nvm i reloaded site

stone urchin
#

hello!

rare gobletBOT
#

Ayo? @stone urchin level 2 !!! lfg

stone urchin
#

someone know a good echo remover ? i try some but they destroy the vocals

wanton jacinth
#

what am i supposed to do if i don't have nvidia gpu
boohooh

low shard
#

idk which u are asking for tho

wanton jacinth
low shard
#

what are u looking for?

low shard
# viral heath so, are there any spaces that can do it?

There are different Text To Speech (TTS) AIs:

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

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

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

With RVC Models:

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

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

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

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

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

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

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

wanton jacinth
rare gobletBOT
#

Ayo? @wanton jacinth level 1 !!! lfg

low shard
# wanton jacinth to create a model and an ai cover in like, any way possible because my attempts ...

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)
  • Lightning.ai (Kinda hard, needs login, no issue with web uis or anything, but only free 15 credits monthly):
  • Mainline (UI, No guide as of right now)
  • Applio (UI, No guide as of right now)
#

i personally suggest kaggles for training (making rvc model)

#

while for making an ai cover, aka inference, id suggest either using applio colab or even better ilaria rvc zero which is fasterr

brave garnetBOT
wanton jacinth
low shard
#

yw

spark nebula
#

I'm a complete beginner for this. I was wanting to make an AI Kanye song from a reference track. I have a Mac and can't seem to have any luck finding a good install guide

#

Any help woould be massively appreciated

buoyant pagoda
#

what rvc choose to make an ai covers

proper shale
proper shale
#

honestly the best way to go is to not use local on Mac imo

proper shale
#

nice! if u have any more issues lmk

spark nebula
#

I just wanted something I could run a vocal through and see if I can change it to how I want it to sound without running into limits

proper shale
#

Which is why the cloud is probably better

spark nebula
wide saddle
#

This might be a a really dumb question, but I’m just looking to change my voice for a goofy script read for like 12 people - does my voice have to be singing for the AI to change it?

proper shale
low shard
#

theres no local mac guide

low shard
#

or if u wanna do it cloud

brave garnetBOT
low shard
#

ilaria rvc zero is the fastest for cloud inference

brittle wing
#

uka???

tame mica
#

its for

#

voices with an english/british accent

#

@cold raptor made it ithink

#

oh wait no

low shard
desert sphinx
#

hello im new to rvc and i want to be able to convert both main vocals and adlibs/background vocals of a rnb/hip hop song, what is the best way to do it? is there software that can do it? or is there a good separator that can separate all vocals etc..

fast cipher
#

I have setup all of this and using the index file and pth. everything shows that its working but no audio is coming through my VB-Audio Cable

#

I cant send images but it all shows its working on Mangio-RVC Realtime-GUI

low shard
#

-rt

azure marshBOT
# low shard -rt

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

low shard
#

which is the first link

#

just suggesting better

fast cipher
low shard
#

yw

crude flame
#

What on earth does this error mean: Unable to allocate 39.9 GiB for an array with shape (1920, 2786138) and data type float64

All i was trying to do was preprocess a dataset. I am assuming that error means that the dataset is to large but idk ive never ran into this before.

forest pewter
#

this is my dataset before processing and UVR:

Sample Rate: 22050
Duration: 10.28 minutes
Number of Samples: 13603329
Bits per Sample: 48000
Channels: Mono (1)

and this is after:

Sample Rate: 22050
Duration: 10.19 minutes
Number of Samples: 13486493
Bits per Sample: 44100
Channels: Mono (1)

I trained on 48k with the latter and the model ended up with tons of artifacts, any suggestions?

rare gobletBOT
#

Ayo? @forest pewter level 1 !!! lfg

tropic phoenix
forest pewter
#

Ight ty, but why is the sample rate so low? I recorded on audacity

#

my mic is the MPM4000U so it's not bad at all

odd shale
#

And prolly because your dataset wasn't really 48k

tropic phoenix
#

Rvc downsamples to 16k when processing and then upsamples to your model sr.

odd shale
#

Por favor no postees fragmentos de audio copyrighteado.

hearty kernel
#

how do i make a voice model for the AI with my own voice?

flint solar
odd shale
#

Also, you can record yourself talking for 20-25 mins

#

-rvc

azure marshBOT
hearty kernel
#

where do i find them?

odd shale
odd shale
hearty kernel
#

okay thank you

hearty kernel
odd shale
odd shale
flint solar
brave garnetBOT
azure marshBOT
brittle wing
#

-colab

azure marshBOT
# brittle wing -colab
☁️ Google Colabs
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
brittle wing
#

Facefusion omg whaaaat

hollow sable
#

-rt

azure marshBOT
# hollow sable -rt

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

hollow sable
#

-rt

azure marshBOT
# hollow sable -rt

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

rare gobletBOT
#

Ayo? @hollow sable level 1 !!! lfg

latent geode
#

wher do i find the models?

#

n when i download models why do they not have crep files or anything i can put into rvc

rare gobletBOT
#

Ayo? @latent geode level 1 !!! lfg

brazen spear
#

Uhmm how to fix this error
"'NoneType' object has no attribute 'setdefault'"??

carmine hearth
#

!howtoask

patent trellisBOT
# carmine hearth !howtoask

How To Troubleshoot AIHC_WaitWhat

__**GIVE CONTEXT.**__ 📝
  • Don't simply mention your issue, like "my rvc is not working".
  • Describe the step you are on, what you're trying to do, the RVC you're using, a screenshot, etc.
  • The more context, the better.
__**BE POLITE.**__ <:matsuripray:1159685390156967936>
  • Don't be desperate. You can ping a Helper, but if they ignore, they aren't available/don't know the answer.
  • It's okay if you're frustrated, but don't take it into this server.
  • Don't DM without prior consent.
__**BE PRODUCTIVE.**__ 🤝
  • Don't ask for every little instruction. Put your own effort & test things by yourself.
  • Don't ask to ask.
  • Check if your answer is a Google search away/on our guides website.
If you've changed your mind about your message, feel free to delete/edit it.
low shard
low shard
# latent geode wher do i find the models?

You can search rvc ai voice model at:

if there isnt one, you can:

low shard
#

be sure u arent using a yt tut

#

most ones are old

final sentinel
#

How do I create an RVC model on my smartphone?

#

How do I create an RVC model on my smartphone?

knotty moth
low shard
#

U have to do it on cloud (remote good pc)

#

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)
  • Lightning.ai (Kinda hard, needs login, no issue with web uis or anything, but only free 15 credits monthly):
  • Mainline (UI, No guide as of right now)
  • Applio (UI, No guide as of right now)
#

Personally id suggst kaggle but the interface is kinda hard on mobile but u can do it

storm yoke
#

wait i have a question when i put in my model of vice it echos how do i change that?

tulip root
#

Trying to use so-vits-fork for voice conversion, however when i try to input a file for config path (below model path) it is asking for a JSON file, however all the models provide an index file as opposed to a json file. Is there a step im missing?

crude bolt
#

ofcourse you can also stop the voicechanger , but i think this is not really what you are trying 😉

low shard
tulip root
low shard
tulip root
rare gobletBOT
#

Ayo? @tulip root level 1 !!! lfg

low shard
tulip root
low shard
#

for inference (using model), even a gtx 1650 would be enough, while for training (making model), its recommended atleast n rtx 2060

tulip root
echo parrot
#

-rt

azure marshBOT
# echo parrot -rt

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

celest owl
#

I cant post images here?

#

how the hell am I supposed to get any help

latent kettle
celest owl
#

Nope!

latent kettle
latent kettle
celest owl
latent kettle
celest owl
latent kettle
celest owl
rare gobletBOT
#

Ayo? @celest owl level 1 !!! lfg

celest owl
#

XD

#

I just leveled up

#

now I can post images

warm swan
#

i too have a problem sadly, using the Ilaria RVC
I'm just trying out to make an AI cover of a song with a space marine voice and after clicking on convert it gives me the usual waiting on gpu and gpu available basically
then it runs for about 10 seconds and then it just says Error, no code no nothing just error

#

i have yet to find a guide to locally do ai covers as both guides are for online / cloud and neither have worked so far

knotty moth
azure marshBOT
knotty moth
#

stop using that old garbage from such 1+ year old video

celest owl
tame mica
#

-mainline

#

oh

tame mica
azure marshBOT
# tame mica -applio

Suggestions for @celest owl

🖥️ Local stuff

Credits to Faze Masta and Antasma for compiling these links.

More commands: -audio, -colabs, -overtrain, -realtime, -rvc, -help
warm swan
tame mica
#

replay if u just wanna make covers is more than enough + easy to understand

warm swan
#

can i insert ai models from #1175430844685484042 in there too? just can't see that in the little preview there

rare gobletBOT
#

Ayo? @warm swan level 1 !!! lfg

tame mica
#

ye

warm swan
#

neat thanks

tawny harbor
#

Hello, where can I find the app for make my voice change ?

odd shale
azure marshBOT
# odd shale -rt

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

odd shale
#

I suggest you to read the Deiteris' one

tawny harbor
#

thanks @odd shale

rare gobletBOT
#

Ayo? @tawny harbor level 1 !!! lfg

tawny harbor
# odd shale -rt

can you share me the link to download the app, i dont find it please

odd shale
tawny harbor
odd shale
#

But what do you want to download, RVC or W-Okada?

#

RVC is for voice conversion with audios, not realtime.

#

W-Okada is for realtime

tawny harbor
#

for a homework to add fun

odd shale
tawny harbor
#

I want to download voice ai (rvc) for a homework

odd shale
#

just check the docs.

#

-rvc

azure marshBOT
tawny harbor
tawny harbor
# azure marsh

I don't understand in uvr where i put the voice for the result

warm swan
#

quick question with Pitch detect method do you guys think is best?
(feel free to ping on reply)

ef3515c35b050410736a71ee930b382b.png
tidal trail
warm swan
#

quick extra throw in, Stem method? what is recommendet-

cinder glen
#

hello i just installed applio and watched a tutorial. im having issues with the voice models not showing in the drop down menu for voice model. from what i understand there should be at least some already included with applio. what am i missing? I appreciate the help.

Screenshot_2024-09-18_164424.png
low ocean
#

Hi my volume and res stay at 0 no matter what, what can i do?

odd shale
#

You can use Kaggle instead of Colab.

#

-kaggle

azure marshBOT
odd shale
#

I would recommend you this one.

#

https://rentry.co/RVC-Mainline-Kaggle

How to use RVC Mainline Kaggle

This guide for Mainline Kaggle is an alternative option to the Mainline Colab notebook for training voice models
It is complete and should walk you through every step of the way since Kaggle has a difficult learning curve. However, it will be updated constantly to go over parts that need more cla...

brave garnetBOT
#
RVC Colabs and Spaces

⠀

Local Forks 🖥️

⠀
Mainline RVC
Original project, suggested for advanced users,
by the RVC-Project team.

Applio
Simplified, suggested for all, by the Applio team.

RVC Studio
Simplified, suggested for all, by SayanoAI.

Mangio-RVC
Simplified, may not be supported anymore, by Mangio621.

AICoverGen
Simple yet great way to make covers, by SociallyIneptWeeb.

Replay
From the greators of weights.gg, excellent product for everyone.
⠀

Made with love by Ilaria 💖
plucky socket
#

hiiii ai hub, so this might not be the best channel to ask this, but i've seen a lot of models based on the drums of songs, that i assume u can run another drum track through and then it sounds the drums of the trained song. has anybody done this with guitar tones? there's a guitarist i like whose tone i REALLY like and idk i'm just theorizing what would happen if i trained a model of a bunch of his guitar playing

narrow sentinel
#

can someone send a tts rvc

daring raft
#

hey, i reached e300 with training then stoped to save it on drive then i started again and it started from epoch 1? how can i solve this, what happens now?

sonic agate
#

did you save D and G files?

daring raft
#

yes

daring raft
brazen spear
#

Uhmm i use the rvc aicover maker and it say it been limit in google collab and if i want to continue i have to pay and this program say can upgrade by gladio deploy i dont know how to do gladio deploy can some one show me the tutorial?? Sorry for bad eng

knotty moth
# daring raft yes

don't forget to load the preprocessed and the training progress to resume

wanton jacinth
#

should I do something about no graphs showing on the TensorBoard part?

brave garnetBOT
terse nymph
#

is ilaria mainline on colab busted? hasnt been working for awhile and i ran out of gpu time on huggingface

low shard
#

Either wait for zerogou time or use the applio colab

terse nymph
#

Gotcha thanks!

low shard
low shard
brazen spear
rare gobletBOT
#

Ayo? @brazen spear level 1 !!! lfg

low shard
brazen spear
#

I get the gpu warning

low shard
#

Either use an alt google acc

#

Or use -kaggle that gives 30 hours free weekly but it's harder

#

-kaggle

azure marshBOT
brazen spear
#

Ok but daily is 1 perday right and it refill??

low shard
#

I'd personally suggest -kaggle if u want more time as they give 30 hours weekly for free of better GPUs, but it's harder and requires phone number so he choice

brazen spear
low shard
#

Yw

low shard
# narrow sentinel can someone send a tts rvc

There are different Text To Speech (TTS) AIs:

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

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

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

With RVC Models:

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

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

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

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

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

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

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

torn swift
#

Hello, are there any guides to separate 2 voices? I have a song here it's a duet and i want to separate the female and male vocal. Thank you

mystic meadow
#

does anyone know how to remove echos and reverb in the vocals, driving me crazy lol

delicate wedge
#

I want to add more voicelines to a existing rvc model, but don't know how. Also I'm using this rvc model maker.

knotty moth
delicate wedge
brisk quarry
#

Is there a ranking of gpus with their average it/s

knotty moth
delicate wedge
#

Same quality.

knotty moth
# delicate wedge Same quality.

from different source? that would prob be less consistent, or how about train the whole "new voicelines" and compare it to the old?

brisk quarry
#

Balls

summer dirge
#

who can send me a good model of a female voice? Please

rare gobletBOT
#

Ayo? @summer dirge level 1 !!! lfg

brisk quarry
dull crypt
#

hello is there a way to instead of having to add an audio file for the pth file to make it say what the voice says is there a section on mangio-rvc-fork where i can just use the .pth's voice to say something by itself without it like combining with mine sort of like eleven labs where it just uses the voice and you type in what it says.
so basically just like can i use the pth file by itself without having to add my voice and where can i do it pls hlep

brisk quarry
#

Ok

lone iron
#

Bro i need help

#

How to fix module not found error pyngrok not found?

brittle wing
#

How much index is good 0.7 or 0.8?

rare gobletBOT
#

Ayo? @brittle wing level 5 !!! lfg

low shard
lone iron
low shard
#

There are different Text To Speech (TTS) AIs:

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

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

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

With RVC Models:

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

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

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

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

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

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

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

lone iron
#

And how to fix that ?

low shard
#

they are 99% outdated

#

what are u tryna do?

lone iron
#

Where is the new one

low shard
#

inference (use models) or train (make models)?

lone iron
#

Im trying to my fav youtuber voice

low shard
#

also what vid are u looking at cus im pretty sure its old

lone iron
#

But keep pyngrok not found

low shard
rare gobletBOT
#

Ayo? @lone iron level 1 !!! lfg

lone iron
low shard
#

or using the model for pre recorded audios, like ai covers

lone iron
#

Yea for game tho

lone iron
low shard
#

you can use it for pre recorded audios tho

lone iron
#

But how to fix pyngrok not found

low shard
lone iron
low shard
# low shard you can use it for pre recorded audios tho

As you can't locally do it on ur phone, i can suggest you Ilaria RVC Zero which is a cloud (running it on a remote good pc) option to do ai covers

Google Docs
Ilaria RVC Zero

Table Of Contents Introduction (with website link) Model Loader (Download & Upload) Inference (use RVC AI Voice Models) Ilaria TTS Settings (Inference) Vocal Separator (UVR) Troubleshooting “No gpu is available for you for 60s” Introduction (with website link) Ilaria RVC Zero, is an RVC (Re...

rare gobletBOT
#

Ayo? @low shard level 104 !!! lfg

low shard
lone iron
#

Did i can send link here or nah?

lone iron
low shard
low shard
lone iron
#

ay sure

lone iron
low shard
#

also it looks like its a copy of ur colab @glad zealot

glad zealot
low shard
low shard
lone iron
glad zealot
#

XD

low shard
#

certain youtubers love views lol

lone iron
#

But its indonesian