#Not powering in
1 messages · Page 1 of 1 (latest)
I don't quite get it, screenshot pls
Elaborate:
- your pc gpu
- what guide/ kaggle link are you using
- what you want to do
this
The perf doesn't read up, its stuck at 0 and its started alreadyb
Great that's wokada deiteris fork
Elaborate:
- your pc gpu
- what you want to do
- a screenshot of the issue and what the problem is
I uploaded the voice model then I was starting the voice changer and it changed but it wasn't reading
wait, you don't have image perms?
What does that mean??
permission to send images
and it changed but it wasn't reading
Elaborate this part
also you still haven't told me your pc gpu, I'm asking since Kaggle is a cloud (remote good pc) for people with a bad pc, and your gpu could be good enough to run it locally
Yes sure send images
8GB Vram and 4060 Rtx
My P.C don't run it well enough
The kaggle ran better before it developed the issue
your pc gpu actually would be able to run it pretty well, you sure you tried the wokada deiteris fork locally?
or did you try it but didn't work well in a demanding game?
It doesn't run smoothly or is it because I didn't train the model correctly ?
wdym smoothly? robotic or lagging?
and also are u using it in pair with a game?
Doesn't sound the model
Yes
What games? Might be it's a demanding game and you didn't set minimum settings
Its not the game even when I record on the fork it doesn't sound the model and its call of duty
You mean it doesn't sound like the model?
That depends on the settings and model
Yes it doesnt
You sure you tried the Wokada deiteris fork locally?
Do you still have it so I can check the settings
I still have it but I'm sleeping out tonight not with the P.C
Will you help me check out the model?
Oh alright
So you sure you wanna use kaggle instead?
Yes its 15gb dedicated
Recently Ngrok had some issues for some users, could you try if it opens now?
Might be their servers are having issues
Everything seems to work till you open the Ngrok link and you get a blank page, right?
yeah ofc dw
I finally got it to work
It doesn’t sound like the model and it sounds so much like a robot
Are you using it locally? Show a screenshot of the wokada
try another model
Yes I’m using it locally
Show a screenshot to check up your settings
Might be a model issue, but it's better to always check the settings
Tag me to a good model so I can test it because I don’t know if I’m training the model right
FO det: rmvpe, onnx is for non nvidia
Put extra to 2.7, 5 could be higher quality but can cause cutoff issues
Oh is that what I’m having issues with?
Maybe try the ones suggested in the guide https://rentry.co/forkvoicechangerguide#voice-models-to-try-out
Thanks vtarcelia for corrections, Nick088 for contributions. Most technical information comes from deiteris.
Latest Version b2332 from December 2024
RTX 5000 series support is here, but not integrated into w-okada itself, it is a stand-alone release. You can get it from here
Translations (outdate...
The settings are better, if u have cutoff issues set extra to 2.7
Let me try it right now
It’s still the same I think it’s the model
u gotta try out with other models
the settings and program is completely fine
it's just not every model is high quality
Show a screenshot of the whole WOKADA