#✨│ai-help
tranquil cliff
copper pebble
sti cazzi manco il mio
cmq col cazzo che faccio i modelli in mangio crepe
provi a fargli fare una nota alta ed esplodono
cio dei sample con amber in mangio crepe
ti stacchi le orecchie
copper pebble
L'ho scritto per trovare una soluzione
ovvio ovvio
static garden
old one is working fine for me and its pretty easy
copper pebble
L'ho scritto per trovare una soluzione
ma a quale cella si dovrebbe bloccare??
finite galleon
tranquil cliff
Non si blocca ma esce strano il pitch
tranquil cliff
Aspe che aggiusto
copper pebble
infering works
for ilaria
dunno abt the others spaces
but since ilaria's one is the best one
we must not worry abt inferences
tranquil cliff
@copper pebble ho usato il tuo modello
copper pebble
<@1065579511183310939> ho usato il tuo modello
quale??
aaa
ajahhaha
pensavo quelli vocali hahha
vabenissimo btw
tranquil cliff
Lol
astral jungle
Yo
copper pebble
What do you often use to train models?
rvc disconnected
brittle wing
What is the name of the application where I can see the results while training a model?
modern oxide
tensorboard
azure marshBOT
This interaction has expired, use the command
/guides realtimeif you wish to see it again.
woven plover
guys why my cmd says pipelineis not initialized?
rare gobletBOT
Ayo? @woven plover level 1 !!! 
sand iris
guys why my cmd says pipelineis not initialized?
#🔍│help-w-okada pinned guides
azure marshBOT
RVC Guides (How to Make AI Cover)
charred jewel
Well then you’re doing something wrong.
brittle wing
im getting a traceback error when i train model help?
honest pond
im getting a traceback error when i train model help?
send the error or screnshoot
bold plover
AttributeError: 'NoneType' object has no attribute 'tobytes' i keep getting this error whenever i make ai covers
brittle wing
-colab
azure marshBOT
neon valve
Anyone has experience with Ultimate Vocal Remover? What is the best setting to get rid of backing vocal?
proud elbow
Anyone has experience with Ultimate Vocal Remover? What is the best setting to g...
UVR 6HP karaoke, not too perfect but the best so far
neon valve
UVR 6HP karaoke, not too perfect but the best so far
Tried it. Still terrible
honest pond
AttributeError: 'NoneType' object has no attribute 'tobytes' i keep getting this...
try using index or not using the idex
neon valve
UVR 6HP karaoke, not too perfect but the best so far
Maybe the song was too hard? It's "Put your head on my shoulder"
honest pond
AttributeError: 'NoneType' object has no attribute 'tobytes' i keep getting this...
When it happened to me it was the model that did not reach the audio frequency but we don't know how to fix it. At most you could convert it to mp3 with an app or change the sample rate.
bold plover
try using index or not using the idex
thanks ill try that now
proud elbow
Maybe the song was too hard? It's "Put your head on my shoulder"
what kind of that actually, vocal doubling/harmonics or duet? if duet it's not possible
neon valve
what kind of that actually, vocal doubling/harmonics or duet? if duet it's not p...
The result
This is before removing backing vocal
The backing vocal bit left behind really mess with my conversion
Guess I'll give up then
copper pebble
Tried it. Still terrible
did u tried karaoke 5??
proud elbow
This is before removing backing vocal
yeah the "back vocal" part that's nearly as loud as lead vocal may be too hard to separate (could be considered as "duet"/group vocal)
btw 5HP is more of removing harmonics, and it sometimes tears some screaming/high-pitched part on my sample
neon valve
Don't worry. I'm not autistic enough to demand absolute perfection
A bit blemish won't matter
proud elbow
yeah I also had some broken parts on some songs I tried to cover
half cove
robust crane
pls say good setting on gtx1660
deep bridge
where do i get the application that uses these voice models?
copper pebble
pls say good setting on gtx1660
u trying to train locally??
deep bridge
ive seen people do it in real time what application does that?
robust crane
u trying to train locally??
I'm just sitting in discord with friends and the program constantly gives out a trail
rare gobletBOT
Ayo? @robust crane level 1 !!!
copper pebble
I'm just sitting in discord with friends and the program constantly gives out a ...
u using voice changer??
copper pebble
ive seen people do it in real time what application does that?
ohh weare talking about voice changin here
robust crane
u using voice changer??
yes
bronze depot
anyone know any good places i can get rvc models?
i've tried #1159289700490694666 and #🔍│find-models but it's really confusing and i can't find any good models
as it shows the most recent instead of most used or actually good models
bronze depot
try weights.gg
oh alr thanks i've never heard of that before
lol i would but idk where to start
would that be for okada?
rare gobletBOT
Ayo? @bronze depot level 1 !!!
bronze depot
sorry im new here i have know idea what im doing lmao
restive talon
i download a voice model and its says something like there is no "mega"
proper shale
dont download models that have mega download links
mint linden
-colab
azure marshBOT
fierce lava
Hey! I am having issue with installation path. How can i change the directory? Or can i?
While installing torchaudio
sand iris
What's happening ?
fierce lava
I don't have enough disk space on C:/ so i gotta install it on other disk.
sand iris
Oh, how are you installing stuff ?
fierce lava
via CMD pip install -U torch torchaudio --index-url https://download.pytorch.org/whl/cu118 with this.
sand iris
Any reasons you are doing a manual installation ?
I do know multiple ways to install packages to another drive
Just gotta change how you have things setup
fierce lava
How can i do that?
rare gobletBOT
Ayo? @fierce lava level 1 !!!
sand iris
You can unless use one of the "releases" archives that contain all the python dependencies & code, and extract that on the drive you want
...Or if you go manual, setup a python venv on the target drive, which will have the dependencies be installed relative to that venv rather than user-wide on the C: drive
wide parcel
yo im actually really curious
is it possible to automatically train models in applio
like once x model is done move on to y model
and train with same parameters
or am i going to have to mod applio
azure marshBOT
RVC Guides (How to Make AI Cover)
wide parcel
?
vapid gust
What batch size is good for 20min hq dataset
modern oxide
What batch size is good for 20min hq dataset
anything above 8 is good put 12
modern oxide
no
modern oxide
what if I use 16?
won't make much difference
vapid gust
won't make much difference
check ur dms
trim ice
@vapid gust hey, I got that dataset from yesterday mixed, could i ask ya to give it a quick listen to see if it sounds good now? I'm not too familiar with this kinda stuff, so it'd be helpful for feedback / double checking
rare gobletBOT
Ayo? @trim ice level 3 !!!
sand iris
mfw vram
deft river
I'm having trouble making models on a new M3 Mac. Keep getting an Unpicklking error and I think it has to do with the hubert.pt?
rare gobletBOT
Ayo? @deft river level 1 !!!
maiden maple
i have a problem with downloading the model, even i do everything right it gives me:
ModuleNotFoundError Traceback (most recent call last)
<ipython-input-8-981787a46cd5> in <cell line: 4>()
2 #@markdown Link the URL path to the model (Mega, Drive, etc.) and start the code
3
----> 4 from mega import Mega
5 import os
6 import shutil
ModuleNotFoundError: No module named 'mega'
deft river
use gpus:
python3.8 train_nsf_sim_cache_sid_load_pretrain.py -e felicia-apple -sr 40k -f0 1 -bs 1 -te 200 -se 5 -pg pretrained_v2/f0G40k.pth -pd pretrained_v2/f0D40k.pth -l 1 -c 0 -sw 1 -v v2 -li 273
INFO:felicia-apple:{'train': {'log_interval': 273, 'seed': 1234, 'epochs': 20000, 'learning_rate': 0.0001, 'betas': [0.8, 0.99], 'eps': 1e-09, 'batch_size': 1, 'fp16_run': False, 'lr_decay': 0.999875, 'segment_size': 12800, 'init_lr_ratio': 1, 'warmup_epochs': 0, 'c_mel': 45, 'c_kl': 1.0}, 'data': {'max_wav_value': 32768.0, 'sampling_rate': 40000, 'filter_length': 2048, 'hop_length': 400, 'win_length': 2048, 'n_mel_channels': 125, 'mel_fmin': 0.0, 'mel_fmax': None, 'training_files': './logs/felicia-apple/filelist.txt'}, 'model': {'inter_channels': 192, 'hidden_channels': 192, 'filter_channels': 768, 'n_heads': 2, 'n_layers': 6, 'kernel_size': 3, 'p_dropout': 0, 'resblock': '1', 'resblock_kernel_sizes': [3, 7, 11], 'resblock_dilation_sizes': [[1, 3, 5], [1, 3, 5], [1, 3, 5]], 'upsample_rates': [10, 10, 2, 2], 'upsample_initial_channel': 512, 'upsample_kernel_sizes': [16, 16, 4, 4], 'use_spectral_norm': False, 'gin_channels': 256, 'spk_embed_dim': 109}, 'model_dir': './logs/felicia-apple', 'experiment_dir': './logs/felicia-apple', 'save_every_epoch': 5, 'name': 'felicia-apple', 'total_epoch': 200, 'pretrainG': 'pretrained_v2/f0G40k.pth', 'pretrainD': 'pretrained_v2/f0D40k.pth', 'version': 'v2', 'gpus': '0', 'sample_rate': '40k', 'if_f0': 1, 'if_latest': 1, 'save_every_weights': '1', 'if_cache_data_in_gpu': 0}
INFO:torch.distributed.distributed_c10d:Added key: store_based_barrier_key:1 to store for rank: 0
INFO:torch.distributed.distributed_c10d:Rank 0: Completed store-based barrier for key:store_based_barrier_key:1 with 1 nodes.
gin_channels: 256 self.spk_embed_dim: 109
INFO:felicia-apple:loaded pretrained pretrained_v2/f0G40k.pth
Process Process-1:
Traceback (most recent call last):
File "/Users/tyleryoung/Documents/Mangio-RVC-Fork/train_nsf_sim_cache_sid_load_pretrain.py", line 181, in run
utils.latest_checkpoint_path(hps.model_dir, "D_*.pth"), net_d, optim_d
File "/Users/tyleryoung/Documents/Mangio-RVC-Fork/train/utils.py", line 206, in latest_checkpoint_path
x = f_list[-1]
IndexError: list index out of range
During handling of the above exception, another exception occurred:
Traceback (most recent call last):
File "/opt/homebrew/Cellar/python@3.8/3.8.18_1/Frameworks/Python.framework/Versions/3.8/lib/python3.8/multiprocessing/process.py", line 315, in _bootstrap
self.run()
File "/opt/homebrew/Cellar/python@3.8/3.8.18_1/Frameworks/Python.framework/Versions/3.8/lib/python3.8/multiprocessing/process.py", line 108, in run
self._target(*self._args, **self._kwargs)
File "/Users/tyleryoung/Documents/Mangio-RVC-Fork/train_nsf_sim_cache_sid_load_pretrain.py", line 201, in run
torch.load(hps.pretrainG, map_location="cpu")["model"]
File "/opt/homebrew/lib/python3.8/site-packages/torch/serialization.py", line 815, in load
return _legacy_load(opened_file, map_location, pickle_module, **pickle_load_args)
File "/opt/homebrew/lib/python3.8/site-packages/torch/serialization.py", line 1033, in _legacy_load
magic_number = pickle_module.load(f, **pickle_load_args)
_pickle.UnpicklingError: invalid load key, 'v'.```
This is the error I'm gettingif my colab says "Waiting to finish the current execution." dose that mean it crashed?
vapid gust
<@619629784380735567> hey, I got that dataset from yesterday mixed, could i ask ...
Yh send in dms
flint cloud
how do i put the dataset
deft river
So I decided to use the colab and now I'm not sure what the path to the dataset is. I tried /content/temp_dataset/ but that didn't work, despite the fact that's what it says it extracted the files to.
proper shale
how do i put the dataset
you can't train on that space
brittle wing
i really need to train, but i have AMD
so i cant do localy
what online method should i use?
proper shale
So I decided to use the colab and now I'm not sure what the path to the dataset ...
/content/dataset
brittle wing
-colab
azure marshBOT
brittle wing
the disconnected one is trash rn
proper shale
i recommend using RVC Disconnected
brittle wing
doesnt sound good at alll
flint cloud
that space uses cpu only, so you cant train
ah shit
proper shale
flint cloud
i recommend using RVC Disconnected
alr
proper shale
peoplle said it isnt good right now
¯_(ツ)_/¯
proper shale
best one for me currently
brittle wing
best one for me currently
alr
olive olive
for the dataset is it better to make a full dataset with only 1 full clip or alot of clip but short
proper shale
for the dataset is it better to make a full dataset with only 1 full clip or alo...
depends on the size
lol
deft river
I use multiple clips from files because I have the luxury of ripping from video games
olive olive
i have 1h24 full clip
proper shale
if its more than 1 hr you can separate in two
- multiple clips hurted my performance once
...it was 1,2K files but uh
yeah
olive olive
So i don't cut my clip ?
proper shale
no need, basically.
proper shale
suuuure
deft river
I keep getting Operation cancelled MyDrive
Because my instance disconnected
What do I do to completely restart this
sand iris
so i cant do localy
dual boot linux 
brittle wing
dual boot linux <:trolley:1159468147133395025>
true
opal saffron
Hoow do I use it on Android
deft river
OSError: [Errno 125] Operation canceled: 'MyDrive'
I get this when I try to run the Colab now :/
And it keeps trashing my Drive files
rare gobletBOT
Ayo? @deft river level 2 !!!
jolly pilot
-help
deft river
My runtime keeps disconnecting because of some command it says it can't execute in the free tier..?
Using the colab
radiant lichen
Does anyone know why the connection keeps timing out and I'll just start getting random errors when trying to convert a song? it literally only gives me a minute and then I have to restart the whole process again it's not enough time to download....
drowsy pilot
can someone tell me why this is taking so long for a 30 second mp3
deft river
Could someone just tell me how to get this running locally on Apple Silicon? The colab refuses to cooperate with me.
All I'm trying to do is train
When I switch to V2 it just loads with a queue that never loads anything now
I really don't know what's wrong
brave garnetBOT
Made with love by Ilaria 💖
thin harbor
my ai when use a voice is hearng bad
the lower the epoch level is the lower quality the voice is
huh, i was told that the higher the epoch the better it is
then what does the epoch mean
obtuse stream
I used applio how can i check the train? I mean the graph
thin harbor
could anybody send me the voicemoderation file
fierce lance
the lower the epoch level is the lower quality the voice is
where i select epoch
rare gobletBOT
Ayo? @fierce lance level 1 !!!
austere wave
where i select epoch
what do you use to train ?
normally it should be clairly noted
austere wave
advanced setting ?
rare gobletBOT
Ayo? @austere wave level 1 !!!
fierce lance
fierce lance
wich app should i use mate?
austere wave
i recommend rvc disconnect
fierce lance
could u send me a link?
wide parcel
yo once again is there a way to queue models for training?
instead of going one by one
azure marshBOT
fierce lance
https://discord.com/channels/1159260121998827560/1160926856548200531/11609268565...
i need the rvx app
rvc
fierce lance
what to do in colab
austere wave
https://discord.com/channels/1159260121998827560/1160926856548200531/11609268565...
^
the instructions are here
austere wave
follow the tutorial
austere wave
austere wave
i dont understand so much
you just need to run cells
and drag and drop files
it's not too difficult
austere wave
no sry
austere wave
if you just want that i think you can find it in existant models #1159289700490694666 here
fierce lance
oke
rare gobletBOT
Ayo? @fierce lance level 2 !!!
fierce lance
yes
fierce lance
oke
austere wave
or on the server
fierce lance
k
atomic mulch
Whats the difference between rvc and okada?
primal wing
hushed quartz
trained a model of omni-man for 650 epochs and the audio sounds robotic? what should i do???
confused as to why it sounds robotic
this was a test sample i recorded using realtime
for dataset it was only one audio file, that was about 6 minutes long, but doubt that is the issue
green birch
trained a model of omni-man for 650 epochs and the audio sounds robotic? what sh...
i get models that sound like cheap voice changers so im in the same boat i think it has to do with the audio
green birch
change ur inferring settings
i have done
300
400
500
600
800
1000
10,000
all the very same rsults
violet heron
Whats the difference between rvc and okada?
Okada is for changing your voice in real time
hushed quartz
change ur inferring settings
inferring settings? if it isn't related to changing epoch level what settings fall under that
violet heron
RVC is for making it speak/sing using a file you provide
atomic mulch
Okada is for changing your voice in real time
How do you update both rvc and okada
green birch
i was using mandingo to ry to make okoda (rvc)
atomic mulch
RVC is for making it speak/sing using a file you provide
THANK U that clears it up so much!
violet heron
How do you update both rvc and okada
I would just reinstall
atomic mulch
I would just reinstall
delete the old ones or just move theem
violet heron
delete the old ones or just move theem
I would save whatever files you need like your weights/logs folder
And move it to the new installation
atomic mulch
I would save whatever files you need like your weights/logs folder
Thank you so much Timbovill!
rare gobletBOT
Ayo? @atomic mulch level 3 !!!
green birch
i need help from somebody that really knows what they are doing so i know if its how im doing it or the audio but nobody will look at it free or at all it seems so i been stuck for almost 2 weeks
not sure what to do tbh
violet heron
i need help from somebody that really knows what they are doing so i know if its...
Just post what your having a issue with
green birch
Just post what your having a issue with
if i knew what iw as having trouble with id just fix it thats the very issue
i have tried many epoics i have tried more than 1 audio with wav and changed to sound "better" took out silence and other things
put a massive amount of time into it and results are the same reguardless of epocs and audio i use so i have no idea
hushed quartz
i haven't done it to the extent you have but my issue seems to be the same
rare gobletBOT
Ayo? @hushed quartz level 2 !!!
vapid gust
Maybe change training settings
vapid gust
if i knew what iw as having trouble with id just fix it thats the very issue
What’s ur issue again
green birch
Maybe change training settings
i have done 20 different set ups
vapid gust
What’s ur issue
green birch
my issue is i dont know what my issue is
all my traing = the same
vapid gust
blud
green birch
different audo different amounts of epoc all sound like a cheap voice changer (aka blowfish) with the helium
vapid gust
just show me what’s the problem
green birch
again i dont know if its the audio or how im doing it if u wanna see the settings i can show u the last ones 1 sec
vapid gust
yes man just show me what’s ur problem
violet heron
i have tried many epoics i have tried more than 1 audio with wav and changed to ...
Are you sure your converting different models each time
green birch
i have done batch settings from 8-64
green birch
1 moment i can send the setup to you via dm
vapid gust
I just want to see how shit it is
hushed quartz
this was a test sample i recorded using realtime
this is an example since we seem to be having the same issue
vapid gust
is that tts?
hushed quartz
no it's not
vapid gust
@hushed quartz
green birch
I just want to see how shit it is
sent, im gonna eat dinner but lemme know if its the audio or what i should do to make it better (ish) im new to this i donno what im doing wrong
no real audio
hushed quartz
<@519651630421901335>
i used realtime rvc
i spoke, and recorded with audacity
and it generated that
with the model i trained
"this is a test of the model, and if it doesn't sound good i will be very disappointed" is what i said
vapid gust
show me the dataset
green birch
you have to remove silence
and sorta equal audio
rare gobletBOT
Ayo? @green birch level 8 !!!
hushed quartz
you have to remove silence
and sorta equal audio
that isn't the issue
the audio is fairly clean
tulip egret
aw the car flippin
tulip egret
Is there anything wrong here?
green birch
i dont know if its the issue or not i was told you have to do that tho
potent cradle
just mp3 convert it
... afaik these programs onlt read wavs
potent cradle
Then download and convert it yourself
vapid gust
Im on mobile I don’t want to download a 60mb file
hushed quartz
vapid gust
Thank you
hushed quartz
i dont know if its the issue or not i was told you have to do that tho
it isn't i have used datasets with noise and silence and it was fine
tulip egret
Is there anything wrong here? It keeps saying error it's a 1 minute sound
hushed quartz
how many epochs is that model <@519651630421901335>
650 is what i trained to
green birch
im limited to 5 megs
to upload as i dont sub to discord
but im told you need to use wav not mp3
hushed quartz
im limited to 5 megs
to upload as i dont sub to discord
but im told you need t...
i sent mp3 so envy could listen
green birch
im told it needs to have silence removed reverb turned down
tulip egret
Hello?
green birch
i sent mp3 so envy could listen
i can see it i just cant upload them
tulip egret
Is there anything wrong here?Is there anything wrong here?Is there anything wrong here?Is there anything wrong here?Is there anything wrong here?Is there anything wrong here?Is there anything wrong here?
green birch
so thats what i know as of now
hushed quartz
tulip egret
JUST TELL ME IF THERES ANYTHING WRONG ITS A 1 MINUTE SOUND
hushed quartz
okay will do, and for the epoch amount what should be good enough for the 6 mins of data?
hushed quartz
no? why
tulip egret
JUST, FUCKING, TELL, ME, IS, THERE, ANY, FUCKING, THING, WRONG
vapid gust
STFU
tulip egret
I HATE YOU IGNORING PERSON
IM JUST TRYING TO ASK IS THERE ANYTHING WRONG BUT I GET IGNORED, IGNORED AND YOU TELL ME TO SHUT THE FUCK UP, WE HAVE A FUCKING PROBLEM 🤜 🤜
half cove
JUST, FUCKING, TELL, ME, IS, THERE, ANY, FUCKING, THING, WRONG
try using somthing that aint rmvpe
or use ila rvc
tulip egret
I did
green birch
JUST, FUCKING, TELL, ME, IS, THERE, ANY, FUCKING, THING, WRONG
try rvc
try being polite
try accepting ppl may not know the answer
tulip egret
Umm
vapid gust
use tensorbord > scalars tab > type loss/g/total > ignore outliners > smoothing 0.999
green birch
not sure how to use it offline
vapid gust
not sure how to use it offline
just find out or something I train on colab so idk how local works
hushed quartz
use tensorbord > scalars tab > type loss/g/total > ignore outliners > smoothing ...
Alright, this is how I find it correct?
vapid gust
that’s how
u find the correct graph to use in tensorbord
to find tensorboard in local itself idk
u gotta ask someone else
green birch
i dont think it works offline and i dont use colab cuz i have mutli video cards sitting here so training is way faster and right away vs when they feel like letting me
i do the same for the ai art
potent cradle
I recently saw this analogy for what overtraining is (forgot who said so but)
Overtraining is like the AI learning to cheat to look good
potent cradle
But as in good to the program
vapid gust
I’m training a model
green birch
I recently saw this analogy for what overtraining is (forgot who said so but)
i get overtraining is a thing but i dont know how to use it offilne and most ppl wont view the audio so its pointless to ask what im doing wrong half the time
half cove
JUST, FUCKING, TELL, ME, IS, THERE, ANY, FUCKING, THING, WRONG
seems like huggingface isnt working rn
green birch
but i have done over 20 set ups and maybe 200 hours of this as testing all sound identical this is my issue over all
potent cradle
You want to avoid overtraining
vapid gust
but i have done over 20 set ups and maybe 200 hours of this as testing all sound...
that’s probably because ur model is overtraining each time u train
green birch
i know its 300 minimal for hq
green birch
there is no adv i told you a few settings i hae done 300-10,000
150 ok i'll be back in a bit
green birch
it only took me a few hours
vapid gust
ur lucky
green birch
it was more a test than anything
vapid gust
400 epochs takes me 5 hours
half cove
models aint all about epochs its more about dataset quality
green birch
i cant even make a model for me
vapid gust
trust
just try my dataset
I mean won’t change anything unless you got ur own setup patterned
green birch
trust
what did u want the batch size at ?
starting training now* i'll get back in like 10 min cuz imma do dinner to rq
flint cloud
just made a model
how do i put it in a zip
is it the same thing?
(pth+index+dataset)?
violet heron
how do i put it in a zip
If your on windows 10, select the added index and pth, right click, send to…, send to compressed zip
On win 11 you have to click something like show more options
flint cloud
ahh
rare gobletBOT
Ayo? @flint cloud level 2 !!!
flint cloud
If your on windows 10, select the added index and pth, right click, send to…, se...
im on android
violet heron
im on android
Default files app?
violet heron
the app not the .rar file
Oh lol
flint cloud
uh are you guys sure its just the pth and the index? it doesnt work when i try to download the model
im using rvc v2 huggingface edition btw
then why doesnt it download?
alright
how do i upload it to the id?
i already uploaded to
OH WAIT
ITS PROBABLY BECAUSE I MADE IT PRIVATE
Hold on let me adjust somethings
it worked!
lol
but thank you anyway dude!
thanks!
hushed quartz
since these links are invalid where can i find them?
https://cdn.discordapp.com/attachments/1117648692468527174/1117649813153005618/TBI.bat
https://cdn.discordapp.com/attachments/1117648692468527174/1117649813538865162/tbL.bat
they are for running tensorboard locally, as i am training a model locally
thank ya
opal saffron
Can someone send me a link for Android
rare gobletBOT
Ayo? @opal saffron level 1 !!!
wheat copper
what do i put for this? im using a Quadro RTX 4000
hushed quartz
i was able to get it running locally, pink is the current model i'm training
so thanks for the help
rare gobletBOT
Ayo? @hushed quartz level 3 !!!
wheat copper
what do i put for this? im using a Quadro RTX 4000
pls help!
cerulean dew
would anyone be willing to make a model for me based off of the audio i give u?
i have all of the audio i want to use but the site i used to use is no longer free
and i would appreciate it if someone could help me out
oh ok
well its not rlly asking for a full model i already did most of the work lol
thats why i asked here
vapid gust
i was able to get it running locally, pink is the current model i'm training
Woah what’s going on w those other ones
Look at the ones that r greyed out
Is that part of ur pink model
hushed quartz
Is that part of ur pink model
they look like other colors but i'm not sure what those are
hushed quartz
vapid gust
Click the ones u want to see
hushed quartz
vapid gust
those lines in the back ground look a bit weird to me
they shouldn’t be that aggressive
hushed quartz
any suggestions on settings i should fiddle with?
vapid gust
Ur graph might’ve already started to overtrain or it’s about to overtrain
yh but I think it’s smoothed out tho
Fully
tru
oh
still
it’s rising though
wheat copper
is the Quadro RTX 4000 good for training i have that
vapid gust
at like 3.5k steps
u can see the line pushing up
I think it’s going to overtrain very soon
or more aggressively overtrain soon
hushed quartz
350 epochs is what I'm training to
But from what I've tested it still sounds like shit
vapid gust
send a sample
hushed quartz
Give me a moment
vapid gust
Alr
rare gobletBOT
Ayo? @vapid gust level 13 !!!
hushed quartz
it's practically at 350 almost but i'm going to bed shortly since it's late but betting the audio result would be not much different
the example audio is 300 epochs
ballin, the voice could be pitched up since the character has a higher voice but the same quality issue persists if i were to
vapid gust
idk it looks like another person has the same issue
so ig u can could ask the owner of the local
creator
brittle wing
I made a voice model, but sometimes my voice cuts out when I speak.
flint cloud
making another model due to some issues and i wanna sleep so is it safe to stop while configuring the epoch thing?
rare gobletBOT
Ayo? @flint cloud level 3 !!!
violet cipher
@thin marten
brittle wing
`Attempting to load audios\2023-11-12_22-53-13.wav....
Traceback (most recent call last):
File "D:\xerot\Downloads\Applio-RVC-Fork\my_utils.py", line 64, in load_audio
ffmpeg.input(file, threads=0)
AttributeError: module 'ffmpeg' has no attribute 'input'
During handling of the above exception, another exception occurred:
Traceback (most recent call last):
File "D:\xerot\Downloads\Applio-RVC-Fork\infer-web.py", line 470, in vc_single
audio = load_audio(
File "D:\xerot\Downloads\Applio-RVC-Fork\my_utils.py", line 71, in load_audio
raise RuntimeError(f"Failed to load audio: {e}")
RuntimeError: Failed to load audio: module 'ffmpeg' has no attribute 'input'
Traceback (most recent call last):
File "C:\Users\xerot\AppData\Local\Programs\Python\Python39\lib\site-packages\gradio\routes.py", line 437, in run_predict
output = await app.get_blocks().process_api(
File "C:\Users\xerot\AppData\Local\Programs\Python\Python39\lib\site-packages\gradio\blocks.py", line 1349, in process_api
data = self.postprocess_data(fn_index, result["prediction"], state)
File "C:\Users\xerot\AppData\Local\Programs\Python\Python39\lib\site-packages\gradio\blocks.py", line 1283, in postprocess_data
prediction_value = block.postprocess(prediction_value)
File "C:\Users\xerot\AppData\Local\Programs\Python\Python39\lib\site-packages\gradio\components.py", line 2586, in postprocess
file_path = self.audio_to_temp_file(
File "C:\Users\xerot\AppData\Local\Programs\Python\Python39\lib\site-packages\gradio\components.py", line 360, in audio_to_temp_file
temp_dir = Path(dir) / self.hash_bytes(data.tobytes())
AttributeError: 'NoneType' object has no attribute 'tobytes'`
does anyone know how to fix this?
this is from my applio rvc fork btw
covert ridge
cerulean dew
does anyone know why the ai model i made sounds super staticky?
remote maple
does anyone know why the ai model i made sounds super staticky?
probably a poorly built dataset
cerulean dew
poorly built in what way
rare gobletBOT
Ayo? @cerulean dew level 1 !!!
remote maple
not consistent, or one that has bad isolation
cerulean dew
i can send the clips i used but i dont know i feel like i tried doing all i could
remote maple
i can send the clips i used but i dont know i feel like i tried doing all i coul...
send them in dms
glad zealot
`Attempting to load audios\2023-11-12_22-53-13.wav....
Traceback (most recent ca...
not quite sure but i think thats because of either corrupted audio file or you are trying to use something thats not an audio file
brittle wing
oh wait nvmd i fixed it by reinstalling python ffmpeg
rare gobletBOT
Ayo? @brittle wing level 2 !!!
static gyro
Epoch 900 and its still going down haha
brittle wing
How can I solve the problem of sound interruption while talking on my model?
ember crow
how do i enter in a desired gpu for the config
and no
i am not digging into the code to find where the gpu is set and how it's set
brittle wing
Why can't I click on the ones that say (N/A)
pliant crescent
Don't worry guys
I fixed melisa's problem
modern oxide
guys help please
What happened
loud berry
What happened
how to separate lead vocals from backing vocals
loud berry
do the links must be huggingface to submit modeles?
harsh scarabBOT
You're placed in queue. Current position: 4
formal quartz
I cant find the voice model that Im looking for
proud elbow
I cant find the voice model that Im looking for
#1159289738314919936 or make a model by yourself
formal quartz
Hi @proud elbow thanks for the fast reply. How can I create the model by myself ?
loud hollow
sounds like a really stupid question but how do i put models into huggingface 
proud elbow
Hi <@233145644238241793> thanks for the fast reply. How can I create the model ...
look for #1159513888199540817 to train locally using Mangio/Applio or online using RVC disconnected
rare gobletBOT
Ayo? @loud hollow level 1 !!!
ember crow
look for <#1159513888199540817> to train locally using Mangio/Applio or online u...
or use mainline rvc locally
mainline rvc has less problems than applio
formal quartz
@proud elbow I already donwloaded the vocal with uvr. But at the guide doesn't say how to create the model to be upload it to Ilaria RCV
ember crow
<@233145644238241793> I already donwloaded the vocal with uvr. But at the guide...
oh you mean uvr
formal quartz
So what I did is that I separate already the vocal and the instrumental
ember crow
oh
formal quartz
but how to create the model if is not available in #1159289700490694666
rare gobletBOT
Ayo? @formal quartz level 1 !!!
proud elbow
<@233145644238241793> I already donwloaded the vocal with uvr. But at the guide...
the ilaria is for inference only, still you can test your trained model there
loud berry
MVSep-MDX23-Colab.ipynb not working for me as desired
meager quarry
look for <#1159513888199540817> to train locally using Mangio/Applio or online u...
I thought rmvpe was still broken on anything other than illaria rvc?
proud elbow
I thought rmvpe was still broken on anything other than illaria rvc?
do you mean inference rmvpe? my mangio and applio are still fine
meager quarry
do you mean inference rmvpe? my mangio and applio are still fine
I meant rmvpe as in training
vale stream
.index file directory: ./models\obama\added_IVF2872_Flat_nprobe_1_v2.index
loading ./models\obama\obama.pth
gin_channels: 256 self.spk_embed_dim: 109
Exception in Tkinter callback
Traceback (most recent call last):
File "C:\Users\AlluXD\AppData\Local\Programs\Python\Python311\lib\tkinter\__init__.py", line 1921, in __call__
return self.func(*args)
File "C:\Users\AlluXD\AppData\Local\Programs\Python\Python311\lib\site-packages\customtkinter\windows\widgets\core_widget_classes\dropdown_menu.py", line 101, in <lambda>
command=lambda v=value: self._button_callback(v),
File "C:\Users\AlluXD\AppData\Local\Programs\Python\Python311\lib\site-packages\customtkinter\windows\widgets\core_widget_classes\dropdown_menu.py", line 106, in _button_callback
self._command(value)
File "C:\Users\AlluXD\AppData\Local\Programs\Python\Python311\lib\site-packages\customtkinter\windows\widgets\ctk_optionmenu.py", line 385, in _dropdown_callback
self._command(self._current_value)
File "E:\App Files\RVC Gui\RVC-GUI-Windows-pkg\rvcgui.py", line 429, in selected_model
get_vc(pth_file_path, 0)
File "E:\App Files\RVC Gui\RVC-GUI-Windows-pkg\rvcgui.py", line 234, in get_vc
print(net_g.load_state_dict(cpt["weight"], strict=False)) # 不加这一行清不干净, 真奇葩
File "C:\Users\AlluXD\AppData\Local\Programs\Python\Python311\lib\site-packages\torch\nn\modules\module.py", line 2041, in load_state_dict
raise RuntimeError('Error(s) in loading state_dict for {}:\n\t{}'.format(
RuntimeError: Error(s) in loading state_dict for SynthesizerTrnMs256NSFsid:
size mismatch for enc_p.emb_phone.weight: copying a param with shape torch.Size([192, 768]) from checkpoint, the shape in current model is torch.Size([192, 256]).
can someone help me with this?
novel hemlock
Hey! Completely new here, just recently got into AI voice changing software. When attempting to use the colab version I get a 'Runtime Disconnected' error, which states 'Your runtime has been disconnected due to executing code that is disallowed in our free-of-charge tier.' Do I need to pay for something somewhere to get it working?
Typically mine appears to work for a few minutes before this happens
proud elbow
Hey! Completely new here, just recently got into AI voice changing software. Whe...
try RVC disconnected
novel hemlock
try RVC disconnected
Do you mean the link in #1159949278270193734 ?
proud elbow
Do you mean the link in <#1159949278270193734> ?
"RVC Disconnected, by Kit Lemonfoot"
novel hemlock
Cheers, I'll take a look
brittle wing
help?
why arent they allocated correctly :c
ohhh do i need to click on the load dataset thing first
it didnt say that on the guide hm
alright nvm i figuered it out!!
novel hemlock
try RVC disconnected
To clarify, RVC Disconnected services the same purpose as the usual RVC Colab? I just ask as I'm still brand new to this all and trying to figure out what I'm looking at here to get it working
rare gobletBOT
Ayo? @novel hemlock level 1 !!!
brittle wing
To clarify, RVC Disconnected services the same purpose as the usual RVC Colab? I...
rvc disconnected is the most polished project for model training
rare gobletBOT
Ayo? @brittle wing level 2 !!!
brittle wing
im using it rn and its literally never been easier
its much much faster than whatever was available a year ago
its going rlly fast unline the 3 hours it took me to train last time i checked this server out
proud elbow
To clarify, RVC Disconnected services the same purpose as the usual RVC Colab? I...
RVC disconnected doesn't use webUI which is banned by Google for free accounts
brittle wing
RVC disconnected doesn't use webUI which is banned by Google for free accounts
and it doesnt jackhammer the drive too right
with the autosaving thing
novel hemlock
rvc disconnected is the most polished project for model training
Ooh okay awesome, though for me I'm just trying to get an AI voice to work in realtime through a model I've downloaded to use on Discord as my computer isn't powerful enough to do it itself, is this still able to do that?
brittle wing
Ooh okay awesome, though for me I'm just trying to get an AI voice to work in re...
uhhh rvc disconnected is for making models i think
https://docs.google.com/document/d/186_a8d32iYxr1UAEm1LAFQvLu2w3huVIg-KTlTHshg8/edit you should use this instead
proud elbow
Ooh okay awesome, though for me I'm just trying to get an AI voice to work in re...
yeah for training only, though PCs with decent mainstream GPU are capable to do local inference tho
novel hemlock
Ahh awesome thanks, I'll take a look at this!
novel hemlock
yeah for training only, though PCs with decent mainstream GPU are capable to do ...
Mine's definitely not, especially if I'm running a game. I've got a GTX 1060 with 3gb ;-;
But thanks for the help you both, greatly appreciated!
brittle wing
@proud elbow (really sorry for bothering) do you happen to know how to set the epochs count correctly
the guide says "based on your dataset length" but like
what does that mean
my dataset is an hour long
nvm figured it out
formal quartz
How do I solve this?
novel hemlock
https://docs.google.com/document/d/186_a8d32iYxr1UAEm1LAFQvLu2w3huVIg-KTlTHshg8/...
This was only for recording from the looks of it, whereas I was looking for realtime voice changing for use on Discord, but thanks regardless for the help appreciate it ❤️
brittle wing
This was only for recording from the looks of it, whereas I was looking for real...
why didnt you say you wanted realtime lol
well
maybe im just dumb but i thought u wanted an input -> output as mp3 or something
this is what i use
its a great guide
novel hemlock
Ooh okay awesome, though for me I'm just trying to get an AI voice to work in re...
I said here, but is no worries!
Ooh, I give a look, thanks again
brittle wing
I said here, but is no worries!
i missed that sorry :c
novel hemlock
i missed that sorry :c
Is okay no worries!
novel hemlock
https://rentry.co/VoiceChangerGuide
This stuff here is all a guide for getting it to work on my own machine right? I unfortunately needed to use it somehow else like on colab, but that just kept on disconnecting me :/
brittle wing
This stuff here is all a guide for getting it to work on my own machine right? I...
yes
uhhh
doing realtime through colab is kinda messy
i dont actually know any guide that tells you how to
but i remember its extremely experimental
novel hemlock
Yeah it seems so, I've found a guide to get it to work, and even got it to work to a much better degree than I could locally on my machine, but it just keeps on disconnecting me which is a real shame
brittle wing
im lost
kinda
i think
last time i tried this was when the old server was less than 1 mil members
low shard
is there smt i can help you with?
brittle wing
no i think it works
brittle wing
im not sure if im looking at the tensorboard thing correctly
i remember loss/g/total as the thing to look at if i want to check for overtraining
low shard
heres the pic from the official ai hub docs site on how to use the tensorboard
rare gobletBOT
Ayo? @brittle wing level 3 !!!
low shard
ai hub docs site?
https://docs.aihub.wtf/guide-to-create-a-model/tensorboard-rvc this is the site dedicated to the tensorboard (the first part is js installing it locally) https://docs.aihub.wtf/ this is the site and up left corner you can click on english and choose the language like italian ai hub site
low shard
your welcome
lament eagle
https://docs.aihub.wtf/guide-to-create-a-model/tensorboard-rvc this is the site ...
I thought this was Tensorboard 101 tutorial in depth explanation
rare gobletBOT
Ayo? @lament eagle level 12 !!!
lament eagle
lol
ember rapids
this pops up if i try to train a model any idea what i do wrong?
oh i think it works like it should be it was just stuck on epoch 1 for like 15 min
tender basalt
where I can download the voice changer?
obsidian moat
tropic sleet
Where do I download the program to use these voice models?
tropic sleet
local forks
Which one of the "local forks"?
proper shale
it's up for you to decide, but my personal recommendation would be either the mainline RVC or Mangio
shy zealotBOT
I'm sorry, I've tried everything 😭
It seems that I can't find any model with your search, please try another query.
**Method 1 🔎: ** The search accuracy is 100% 💕
Powered by ModelAI.net, Made with Love ❤️
half cove
worthy lantern
why not public url dont show up rn
half cove
hai
crisp plover
Hey, what is a good website or program to use for making singing / rap vocals and than use that in RVC so i can make my model sing / rap the lyrics ?
Would be nice if the vocals are a bit clean and not to much distorted or have alot of reverb etc.
Anyone can help me with this, let me know 🙂
hushed quartz
so ig u can could ask the owner of the local
who is the creator of the local one exactly?
brittle wing
Hi uh
rare gobletBOT
Ayo? @brittle wing level 1 !!!
vapid gust
who is the creator of the local one exactly?
idk but it’ll be hard to contact
I suggest trying colab
I’ll show u an example of my model trained on colab
celest trail
Hi where i can find RVC V2 with gui
molten pecan
Hi where i can find RVC V2 with gui
-rvc
azure marshBOT
Hey, @celest trail!
👇 Here are some useful resources to help you learn how to make ai covers
hushed quartz
idk but it’ll be hard to contact
is it tolerable? doesn't colab require credits
copper pebble
is it tolerable? doesn't colab require credits
no
hushed quartz
there's a way to bypass the disconnecting then?
copper pebble
there's a way to bypass the disconnecting then?
what r u using??
u tried rvc disconnected??
tropic sleet
I downloaded the original RVC. How do I run it?
rare gobletBOT
Ayo? @tropic sleet level 1 !!!
vapid gust
is it tolerable? doesn't colab require credits
if u wanna train for a long time u can go on and buy but free version will work just as fine
vapid gust
there's a way to bypass the disconnecting then?
Autoclifker
hushed quartz
Autoclifker
Can't run Colab locally right
hushed quartz
u tried rvc disconnected??
snow pasture
I literally make all my model on android using rvc disconnected. Only issue is that you can't let the phone screen turn off or be on another app or tab for too long otherwise it'll disconnect meaning you have to keep on the colab and make sure the screen doesn't turn off till complete. (Well that's how I do things)
copper pebble
there's a way to bypass the disconnecting then?
yes
to stop colab runtime disconnect, Ctrl+ Shift + i to open inspector view . Then goto console and paste:
function ClickConnect(){
console.log("Working");
document.querySelector("colab-toolbar-button#connect").click()
}
setInterval(ClickConnect,60000)
try this
@hushed quartz
hushed quartz
alright will do
copper pebble
or just use rvc disconnected it should work
hushed quartz
do you have the link to the model creation colab?
copper pebble
yh just a momnt
hushed quartz
thx
hushed quartz
i had the issue while running regular rvc
rare gobletBOT
Ayo? @hushed quartz level 4 !!!
hushed quartz
to stop colab runtime disconnect, Ctrl+ Shift + i to open inspector view . Then ...
this is considered a disconnect prevention method right?
just clarifying
copper pebble
just clarifying
yes
i u didnt used the "disconnected" colab arleady in the past u can also use it without that
hushed quartz
ah alright
peak tusk
thats why you can select rvc or kits models
cerulean dew
i stg ive tried restarting the runtime several times but this cell just wont work for me
half cove
have you done feature extraction?
cerulean dew
im gonna have to restart the whole model now bc now the feature extraction is broken too from trying to fix it
yes i did try
and it broke
🥲
half cove
cerulean dew
im gonna have to restart the whole thing arent i
hellpppp
everything is going wrong
i literally cant run a single cell
i keep getting this error
cerulean dew
yes but things started going wrong and it said to restart the runtime if that happened and i tried that and everything just started getting worse
and its like when i tried to restart it was still saving things from the last time i opened it
and i couldnt rlly get a proper restart
u dont have to be so condescending abt it
it isnt my fault this program glitches a lot
it worked perfectly fine for me my first attempt at the model
im not stupid
cerulean dew
u i think u need to restart all
r u even gonna tell me what the issue is or just accuse me of not following the guide...
rare gobletBOT
Ayo? @cerulean dew level 2 !!!
copper pebble
r u even gonna tell me what the issue is or just accuse me of not following the ...
never accused u blud
i just wanted to be sure that u followed it
its normal to have problems with colab
i also have a lot of them everyday
after a page refresh the codes can bug
for the autosave
and the cell bug aswell
cerulean dew
what do i do to fix it?
copper pebble
and the cell bug aswell
in that case the best thing to do is restart
cerulean dew
but how tho it keeps stubbornly saving certain things that r not working
delete the cells and reopen the link?
copper pebble
option 1: button above "runtime" > "disconnect and delete runtime"
option : change acc
worn elm
ello how can i start doing a rvc
molten pecan
ello how can i start doing a rvc
-rvc
azure marshBOT
Hey, @worn elm!
👇 Here are some useful resources to help you learn how to make ai covers
loud hollow
what does AttributeError: 'NoneType' object has no attribute 'tobytes' mean?
low shard
when will aod come out lol
last coral
why can i still hear my self and the ai?
frosty eagle
Hi! I have a problem that some input voice doesn't get converted. Is it related to the high pitch of input voice? I am getting paranoid, because I cannot understand whether the problem is in the model or I could tweak some params? I am using RVC models and downloaded different checkpoints.
For some voice it slightly similar, for others it seems like nothing (drake rapping, or frozen let it go high pitch notes)
pallid thicket
Are there any news on whether there will be a method to use RVC without a high end GPU? Ik of the Google way but I mean without a lot of delay or stuttering.
molten pecan
Are there any news on whether there will be a method to use RVC without a high e...
Yeah you can use the huggingface spaces
Here's one
molten pecan
Hi! I have a problem that some input voice doesn't get converted. Is it related ...
What software are you using?
tawny jacinth
hello guys how install on mac? pls help me
pallid thicket
Yeah you can use the huggingface spaces
My bad, I meant real time voice changer
molten pecan
My bad, I meant real time voice changer
Oh try asking on #🔍│help-w-okada
molten pecan
hello guys how install on mac? pls help me
-rvc
azure marshBOT
Hey, @tawny jacinth!
👇 Here are some useful resources to help you learn how to make ai covers
scarlet sapphire
hi there, im running into an error on colab when i copy/install the stuff
do i need to be worried about it? if yes, what exactly do i need to do?
ERROR: pip's dependency resolver does not currently take into account all the packages that are installed. This behaviour is the source of the following dependency conflicts.
torchdata 0.7.0 requires torch==2.1.0, but you have torch 2.0.1 which is incompatible.
torchtext 0.16.0 requires torch==2.1.0, but you have torch 2.0.1 which is incompatible.
torchvision 0.16.0+cu118 requires torch==2.1.0, but you have torch 2.0.1 which is incompatible.
molten pecan
hi there, im running into an error on colab when i copy/install the stuff
do i n...
I don't use Colab but from what I am seeing you need to update "Torch"
Whatever that is
scarlet sapphire
i tried to update that to 2.1.0, but then torchaudio wanted 2.0.1
molten pecan
i tried to update that to 2.1.0, but then torchaudio wanted 2.0.1
That's odd
I don't know what to do about that
Try asking @viscid moss
viscid moss
What's happening?
viscid moss
hi there, im running into an error on colab when i copy/install the stuff
do i n...
What colab exactly?
scarlet sapphire
the realtime-voice-changer
rare gobletBOT
Ayo? @scarlet sapphire level 1 !!!
viscid moss
the realtime-voice-changer
scarlet sapphire
ah, no the okada one
viscid moss
https://colab.research.google.com/github/hinabl/voice-changer-colab/blob/master/...
Try with this version
I think is working
scarlet sapphire
i will try, thanks
lyric fern
Where can I find a RVC colab to use?
little owl
for the rvc v2 disconnected colab, how would i save my models to prevent losing my progress while training? would i have to stop training and run this cell?
viscid moss
for the rvc v2 disconnected colab, how would i save my models to prevent losing ...
stop and run
little owl
stop and run
so stop the training and then run the export model one right?
viscid moss
so stop the training and then run the export model one right?
Yes, but it is only necessary if you no longer want to continue training or if you are going to run out of 6 hours a day of colab.
little owl
Yes, but it is only necessary if you no longer want to continue training or if y...
ok i see, thank you.
stoic rampart
hi does anyone know what ai tool to use to make the backing vocals for ai covers ?
more or less trying to make it sound professional
rare gobletBOT
Ayo? @stoic rampart level 3 !!!
proven hill
chop chop ai person
out of spite i wont help you
loud sage
Im Using Ilaria RVC on HuggingFace and it keeps saying Error when trying to convert. In the logs it says, " AttributeError: 'NoneType' object has no attribute 'tobytes' " Does anyone know whats wrong?
proven hill
Im Using Ilaria RVC on HuggingFace and it keeps saying Error when trying to conv...
its the public one or the duplicated one? i suggest you duplicating the space
should be working now
forest aurora
when trying to train a model, it just doesn't work
After a few failed attempts at normal epoch counts, I set it to only 10, clicked train model, and waited 30 minutes
my vram was occupied and my gpu was doing something, but after 30 minutes it gave me this error
RuntimeError: [C:\actions-runner_work\pytorch\pytorch\builder\windows\pytorch\third_party\gloo\gloo\transport\uv\unbound_buffer.cc:67] Timed out waiting 1800000ms for recv operation to complete
I'm going to reinstall, but is there anything else I can do?
loud sage
its the public one or the duplicated one? i suggest you duplicating the space
It is the duplicated space, i have no idea why it's not working
It was working perfectly fine the other day
vapid gust
do you have it duplicated?
Yes
loud sage
I duplicated it again and no worki
proven hill
It is the duplicated space, i have no idea why it's not working
mh, try restarting it
seems dumb but istg 99.9999% of the time it will be resolved
forest aurora
when trying to train a model, it just doesn't work
After a few failed attempts a...
I reinstalled, it gives me this error which I also saw before
Process Process-2:
Traceback (most recent call last):
File "C:\Users\jc\Downloads\Applio-RVC-Fork\lib\infer\modules\train\train.py", line 280, in run
utils.latest_checkpoint_path(hps.model_dir, "D_*.pth"), net_d, optim_d
File "C:\Users\jc\Downloads\Applio-RVC-Fork\lib\infer\infer_libs\train\utils.py", line 212, in latest_checkpoint_path
x = f_list[-1]
IndexError: list index out of rangeDuring handling of the above exception, another exception occurred:
Traceback (most recent call last):
File "multiprocessing\process.py", line 315, in _bootstrap
File "multiprocessing\process.py", line 108, in run
File "C:\Users\jc\Downloads\Applio-RVC-Fork\lib\infer\modules\train\train.py", line 317, in run
logger.info(
UnboundLocalError: local variable 'logger' referenced before assignment
INFO:Laplace:<All keys matched successfully>
INFO:Laplace:Loaded pretrained assets/pretrained_v2/f0D40k.pth
INFO:Laplace:<All keys matched successfully>
raw wadi
Hello!
My voice changer isn't booting, keeps giving me this message "[VCClient] wait web server" and goes forever counting but doesn't start.
I tried the ip version, didn't connect as well.
merry spindle
why does the ai voice go insane whenever i input these?
oes anybody have an answer
low shard
W I NEED IT ASAP
ik but it will make everyone more reassure abt overtraining as alot of people are worried abt epochs
instead it automatically stops before overtraining so like its easier for everyone
still can help to do it in less time, sometimes people leave training afk
Yea
trim slate
yo when i try to open the .bat file it just flickers with the command prompt and shuts. anyone know how to fix?
could you expand on W10?
glad zealot
Windows 10
glad zealot
There should be something on top that shows what folder you are in
trim slate
hang on im redownloading the file
glad zealot
You can just erase whatever is in there and type cmd
trim slate
oh ok lol
ill do that when the winwar downloads
and that should open it without the flicker?
rare gobletBOT
Ayo? @trim slate level 1 !!!
trim slate
then it opens the actual voice modulator
ah
woops
my fault
didnt know the difference
So i would change the top of the file to cmd
and put in the start_http.bat file and it should work?
after the command prompt opens
alr
thank
bright vine
can someone @ me or send me where is the guide for making your own models pls 😭
bright vine
tysmmm
modern oxide
probably yes, anyhow you already have 40 k steps, that's more than enough for a model
little owl
probably yes, anyhow you already have 40 k steps, that's more than enough for a ...
ah ok thank you
rare gobletBOT
Ayo? @little owl level 2 !!!
proper shale
hm, I'd say no because D is still going down
Keep going
If D starts stagnating then I think it's probably overtraining
merry spindle
oes anybody have an answer
anybody
vapid gust
ur dataset doesn’t look too good
vale stream
Is there a way to stop getting the disallowed notice on Google colab? I use aicovergen no web ui on colab
rare gobletBOT
Ayo? @vale stream level 1 !!!
hot lagoon
what is the best batch size for a 1 hour dataset like 4 or 6 ?
copper grotto
ayo guys what happened to this space? it says "runtime error" when i run it
bold siren
I'm trying to fuse 2 models together but doesn't seem to work.. anyone knows what i'm doing wrong?
modern oxide
what is the best batch size for a 1 hour dataset like 4 or 6 ?
why did you even made 1 hour dataset
crisp plover
I'm trying to fuse 2 models together but doesn't seem to work.. anyone knows wha...
What happens when you fuse 2 models together thats a new one for me 😮 ?
It will just change the voice sound a bit or ?
proper shale
I'm trying to fuse 2 models together but doesn't seem to work.. anyone knows wha...
name has spaces
crisp plover
name has spaces
What happends whe nyou fuse 2 models together ? will it just sound a bit different or something ?
or can you like enhance one model to sound better on some aspects ?
proper shale
What happends whe nyou fuse 2 models together ? will it just sound a bit differe...
Will make a mix of two voices
crisp plover
when you fuse them
crisp plover
Will make a mix of two voices
Mhhh never tried that one before
crisp plover
Check pins for one, AICG with WebUI
Do you know how long that will take more or less mixing/fusing 2 .pth files ?
do you have to train that or something ?
proper shale
Do you know how long that will take more or less mixing/fusing 2 .pth files ?
Should be a quick process, but I'm not entirely sure.
crisp plover
Alright thanks
crisp plover
I'm trying to fuse 2 models together but doesn't seem to work.. anyone knows wha...
i am trying this as well but getting the same error
No spaces in names
crisp plover
name has spaces
you are talking about the .pth file right ?
or it cant have any spaces anywhere ?
solar plover
what would be a good cpu to pair with a rtx 3060 (i am not using a cpu for rendering i just need a new cpu)
crisp plover
name has spaces
I dont think this error has to do with any spaces ?
i have made 100% sure there are no spaces
do you know what it could be ?
crisp plover
what would be a good cpu to pair with a rtx 3060 (i am not using a cpu for rende...
I bought the intel 13th gen
pricey but ok for me
solar plover
wild
i've been using the ryzen 5 3500 with an rtx 3060 on a hp prebuilt
it's a bit balls and the psu is probably gonna fucking explode soon cause it's not enough power to the board and gpu
(recc to have 550w but there's only 400 in the machine rn)
tawny jacinth
is it possible to run in the decor using a macbook on an intel processor?
tender basalt
what is the best voice model to prank someone?
molten pecan
what is the best voice model to prank someone?
The kalomaze one
shy zealotBOT
**Method 1 🔎: ** The search accuracy is 100% 💕
Powered by ModelAI.net, Made with Love ❤️
tender basalt
Wait no not the one screaming
too late
proper shale
do you know what it could be ?
probably diff epoch count or smth
maybe that causes the problem
crisp plover
probably diff epoch count or smth
Not sure as well, but it didnt work no matter what i tried so i gave up on that option for now.
meager elk
how do i connect this voice changer to discord?
proper shale
virtual audio cable
solar plover
personally vb audio cable is a pretty good one to use
proper shale
^^
loud berry
anyone who knows if there is x-minus.pro bot?
proper shale
probably not.
elfin bay
-rvc
azure marshBOT
RVC Guides (How to Make AI Cover)
rare gobletBOT
Ayo? @elfin bay level 1 !!!
modern oxide
What's the link for the rvs
elfin bay
Did the old link change? What's the new one most people use
solid shard
Yo did Clebersla's space rvc stop working to you too?
azure marshBOT
RVC Guides (How to Make AI Cover)
quasi dagger
-overtrain
azure marshBOT
Hey, @brittle wing!
👇 Here are some resources to help you identify if your model is overtraining
All-In-One Guide on how to make a good model
This guide explains how the D and G files works and much more: https://rentry.org/RVC_making-models
Credits: LUSBERT 
Automated Overtraining Detection (AOD)
Will be available soon in #1159513888199540817
Credits: grvyscale
quasi dagger
Missing AOD there
brazen token
how to use models?
mellow flax
Can anyone help me? I installed applio locally and want to use uvr, but 10/15 minutes after entering the data, literally nothing happens in the console or in webgui. any solutions?
cyan tree
rvc doesnt work
Traceback (most recent call last):
File "site.py", line 169, in addpackage
File "<string>", line 1, in <module>
File "<frozen importlib._bootstrap>", line 562, in module_from_spec
AttributeError: 'NoneType' object has no attribute 'loader'
Remainder of file ignored
Error processing line 1 of D:\RVC-GUI-pkg\runtime\lib\site-packages\matplotlib-3.6.2-py3.9-nspkg.pth:
Traceback (most recent call last):
File "site.py", line 169, in addpackage
File "<string>", line 1, in <module>
File "<frozen importlib._bootstrap>", line 562, in module_from_spec
AttributeError: 'NoneType' object has no attribute 'loader'
Remainder of file ignored
Error processing line 7 of D:\RVC-GUI-pkg\runtime\lib\site-packages\pywin32.pth:
Traceback (most recent call last):
File "site.py", line 169, in addpackage
File "<string>", line 1, in <module>
ModuleNotFoundError: No module named 'pywin32_bootstrap'
Remainder of file ignored
Traceback (most recent call last):
File "D:\RVC-GUI-pkg\rvcgui.py", line 3, in <module>
from tkinter import filedialog
ModuleNotFoundError: No module named 'tkinter'
everytime i try andd run it it throws this up
cerulean dew
I'm fr gonna cry if this site tells me one more time that my audio doesn't have any features to extract
Which makes it so i can't index train
So i can't make the model
cyan tree
im confusedd how ddo i fix my error?
iddddd dont know how to do c++
or whatever this shit is
ive tried installing c++
no fix
cerulean dew
i keep getting this message no matter how many times i try to reset the runtime and get the feature extraction to run properly
ive literally tried deleting every single cell, deleting my rvcDisconnected folder completely and starting completely from scratch 5 times and its still broken
this site literally does not work
if someone has gotten it to work maybe join a vc and help me see why it keeps going wrong?
proper shale
i keep getting this message no matter how many times i try to reset the runtime ...
codec mismatch on the dataset. where did you export it to?