/sdg/ - Stable Diffusion General

actually nevermind it keeps giving me fucking nudes instead of clothes

Attached: 1651354390763.png (512x512, 465.49K)

relative of

Attached: 00037-1994670171-teen_girl,_large_grey_eyes,_black_hair,_clean_cel_shaded_vector_art,_shutterstock,_behance_hd_by_lois_van_baarle,_artgerm,_helen.png (512x512, 270.82K)

Attached: 01389-4206653366-Fantasy_portrait_of_monk_fighter,_wearing_intricate_thin_flowing_robes,_orrery_room_over_shoulder,_rembrandt_lighting,_by_Greg_R.png (512x640, 432.14K)

Rebecca Ferguson from X Factor and Rebecca Ferguson from MI are fused together in a simple prompt
>Rebecca Ferguson

Absolute kino

Attached: hitler portrait.png (512x512, 452.35K)

Attached: 20220905-002911.png (1080x1081, 2.07M)

What is stable diffusion?

its diffusion, just stable

I learn!

was afraid of high de-noising, since I didn't want the inside to be varied, but changing de-noise from 0.35 to 0.8 went a long way
idk if changing from euler_a to euler mattered, but i wanted more stable progression, if anything, since I upped steps from 16 to 40

Attached: Capture1.png (1057x878, 675.64K)

>Swedish actress Rebecca Ferguson
bros...

Attached: grid-0000-70868777.png (1536x1536, 3.32M)

here's the paper:
arxiv.org/pdf/2112.10752

Attached: grid-0246.jpg (1536x2112, 398.46K)

Stable Diffusion is a machine learning model developed by Stability AI to generate digital images from natural language descriptions. The model can be used for other tasks too, like generating image-to-image translations guided by a text prompt.

Attached: 1642703265620.jpg (2048x1536, 646.02K)

Attached: sgnloi.png (1024x1024, 1.55M)

set CFG to 15

It seems to be heavily biased for nudes. Ya know, because most images on the internet are of naked people.

Kinda useless

That's her alright

Attached: 1510002145636.png (720x960, 808.29K)

Attached: 00047-2683834569-teen_girl,_wide_black_eyes,_sapphire_hair,_dark_skin,_clean_cel_shaded_vector_art,_shutterstock,_behance_hd_by_lois_van_baarle,_.png (512x512, 252.46K)

that's a good artstyle user

Attached: grid-0245.jpg (1536x2112, 530.33K)

Why’d it niggafy her?

my weeb was showing and i thought bottom left was rin and top middle was asuka from thumbnail alone before i clicked the full image

can stable diffusion run on TPU with any speed increase?

google "rebecca ferguson x factor"

>Rebecca Louisa Ferguson Sundström

Attached: grid-0000-3140004527.png (1536x1536, 3.39M)

>rentry.org/Stable-Diffusion-Training
so i pretty much have to do this in linux correct?

hello yes how do i make ai booba

Thank you anons who pointed out what files to change to increase batch count!

Going from being limited to 16 to 100 goes a long way

Attached: 01531-3458044146-Fantasy_portrait_of_old_sage,_wearing_intricate_flowing_clothing,_indoors,_strong_lighting,_triangular_glass_window_behind,_cand.png (512x640, 491.47K)

>go into settings
>change grid file format to "jpg"
>save
>generate some prompts
>no grid is being generated in my grid folder
>seethe
>settings
>change jpg to jpeg
>works now
>mfw

Attached: 1424549594156.png (512x512, 410.82K)

>ui-config.json
in voldy root folder for anyone wondering

Attached: 1443239382662.png (512x512, 367.7K)

Extremely, EXTREMELY questionable content warning, but this guy might go good in your prompts: pixiv.net/en/users/1489704

Goddamnit.

Still don’t know why it thinks Juno Temple is black.