Stable diffusion thread

doogyhatts

Arch-Supremacy Member
Joined
Feb 13, 2018
Messages
12,267
Reaction score
3,348
At some point, we will have to train our own tensor models.
That means koping Twmm,Krmm,Jpmm,Hkmm,Atb pictures from their IG/FB.
I haven't figured that one out yet but will start learning SD once I finish my current NFT project.
For now, I will stick to the recommended models.

I suspect there is a lot of advert money can be made on YT from doing all these Kym stuff.
 
Last edited:

DulanMax

High Supremacy Member
Joined
Jul 24, 2020
Messages
27,711
Reaction score
25,132
At some point, we will have to train our own tensor models.
That means koping Twmm/Krmm/Atb pictures from their IG/FB.
I haven't figured that one out yet but will start learning SD once I finish my current NFT project.
For now, I will stick to the recommended models.

I suspect there is a lot of advert money can be made on YT from doing all these Kym stuff.
Maybe create NSFW pics for Onlyfans...

See which simps horny enough to pay money for AI generated NSFW pics

:s13:
 

glarerder

Great Supremacy Member
Joined
Oct 6, 2018
Messages
59,359
Reaction score
22,476
At some point, we will have to train our own tensor models.
That means koping Twmm/Krmm/Atb pictures from their IG/FB.
I haven't figured that one out yet but will start learning SD once I finish my current NFT project.
For now, I will stick to the recommended models.

I suspect there is a lot of advert money can be made on YT from doing all these Kym stuff.
can use 3d printing to print the body and apply ai to give it intelligence. use robotics to give it motion capability
 

doogyhatts

Arch-Supremacy Member
Joined
Feb 13, 2018
Messages
12,267
Reaction score
3,348
Maybe create NSFW pics for Onlyfans...

See which simps horny enough to pay money for AI generated NSFW pics

:s13:
Taiwanese guys can do, we cannot since our legal system forbids porn production.

can use 3d printing to print the body and apply ai to give it intelligence. use robotics to give it motion capability
This one is for Japan scientists & engineers to produce. We won't be able to catch up in robotics hardware tech ourselves.
 

zheng

Banned
Joined
Dec 20, 2001
Messages
52,689
Reaction score
48,715
I just realised that my legacy gpu cannot run stable diffusion. 🤣
yah need a gpu with 8gb vram but 12gb is recommended... think can configure to run on cpu only... if not just pay for google colab
 

zheng

Banned
Joined
Dec 20, 2001
Messages
52,689
Reaction score
48,715
At some point, we will have to train our own tensor models.
That means koping Twmm,Krmm,Jpmm,Hkmm,Atb pictures from their IG/FB.
I haven't figured that one out yet but will start learning SD once I finish my current NFT project.
For now, I will stick to the recommended models.

I suspect there is a lot of advert money can be made on YT from doing all these Kym stuff.
previously chilloutmix model with jp,kr,tw lora... are train base on ig pics... but the creator take down already... probably due to some controversies... model is now transfer to civitai and re-release as chilloutmix-ni
 
Last edited:

x1243x

Arch-Supremacy Member
Joined
Oct 29, 2004
Messages
11,687
Reaction score
3,596
yah need a gpu with 8gb vram but 12gb is recommended... think can configure to run on cpu only... if not just pay for google colab
interesting.. is it worth paying extra for the gpu to play with stable diffusion?
 

zheng

Banned
Joined
Dec 20, 2001
Messages
52,689
Reaction score
48,715
interesting.. is it worth paying extra for the gpu to play with stable diffusion?
for me work related so worth it... just buy a 2nd hand 2060 8gb or 3060 12gb good enough already
 

x1243x

Arch-Supremacy Member
Joined
Oct 29, 2004
Messages
11,687
Reaction score
3,596
for me work related so worth it... just buy a 2nd hand 2060 8gb or 3060 12gb good enough already
ah ok.. I'm still in the market to upgrade.. may add this to the list of considerations. but maybe only 8gb
 

x1243x

Arch-Supremacy Member
Joined
Oct 29, 2004
Messages
11,687
Reaction score
3,596
i only have few pics of my ideal type, like that also can right?

transfer learning
ask ts.. I'm still a noob using the online one.. but I'm guessing the more you feed it the more accurate it'll be.. there's a reason why you see the designs that are popular are quite similar.. all using the same template.
 

sales69

Arch-Supremacy Member
Joined
Jun 22, 2003
Messages
11,639
Reaction score
7,311
What is this about arh?

Anyway I vote for Miss Motherland. 😍 (hope this is okay, don't want to derail your thread)
 

endo5x5

Master Member
Joined
Jan 3, 2018
Messages
4,468
Reaction score
2,619
if moi feed in a hundred images of moi favourite AV stars (from Jpn, HK and TW) as reference, will stable diffusion help create the perfect AV star of my dreams? :(
 
Important Forum Advisory Note
This forum is moderated by volunteer moderators who will react only to members' feedback on posts. Moderators are not employees or representatives of HWZ. Forum members and moderators are responsible for their own posts.

Please refer to our Community Guidelines and Standards, Terms of Service and Member T&Cs for more information.
Top