/sdg/ - Stable Diffusion General

Old thread: >Starter guides:
Google stable diffusion automatic webui (user guides are outdated)

>Krita Plug-ins
github.com/sddebz/stable-diffusion-krita-plugin
github.com/nousr/koi

>Linux
github.com/hlky/stable-diffusion/wiki/Linux-Installation

>AMD
rentry.org/ayymd-stable-diffustion-v1_4-guide
rentry.org/sdamd - Updated AMD guide

>Training
rentry.org/Stable-Diffusion-Training

>Compute Platforms with free GPUs
- Colab
- PaperSpace

>Colab (HLKY)
colab.research.google.com/drive/1AfAmwLMd_Vx33O9IwY2TmO9wKZ8ABRRa

>In-depth usage information
wiki.installgentoo.com/wiki/Stable_Diffusion - Any Forums Wiki with tips
github.com/Maks-s/sd-akashic - Stable Diffusion Akashic Records

>Tools &Resources
pharmapsychotic.com/tools.html - Big index
sgreens.notion.site/sgreens/4ca6f4e229e24da6845b6d49e6b08ae7 - Artist styles
lexica.art/ | krea.ai/ - Prompt sharing sites

>Debugging
rom1504.github.io/clip-retrieval/
laion-aesthetic.datasette.io/laion-aesthetic-6pls/images
replicate.com/methexis-inc/img2prompt - SD Interrogator

>SaaS sites
huggingface.co/spaces/stabilityai/stable-diffusion - Free txt2img
huggingface.co/spaces/huggingface/diffuse-the-rest - Free img2img
beta.dreamstudio.ai - Free txt2img, New accounts get 200 generations
pornpen.ai - Free txt2img, NSFW focused
dezgo.com - Free txt2img
huggingface.co/naclbit/trinart_stable_diffusion - txt2img specializing in animu

>Related
/h/ /sdhb/: /vg/ /aids/: /mlp/ /ppp/: Official discord: discord.gg/stablediffusion
NSFW discord: discord.gg/2ZhBzDGf8u
Tech support Matrix: matrix.to/#/#info:conlaboro.xyz
Stability.ai updates: twitter.com/EMostaque

Attached: 1608989846331.jpg (1536x1536, 434.58K)

>d&d fantasy winter pinetree forest background (cabin). by Gabriele Dell'Otto + Jeff Easley + Larry Elmore + William O'Connor + Tyler Jacobson + Gerald Brom + Heonhwa Choe + E.M. Gist + artstation + deviantart, soft studio lighting, ultra realistic
Man those are some strong Winter Wonderland vectors.

Attached: 00092-2051939797-d&d_fantasy_winter_screaming_horror_apocalypse__pinetree_forest_background_(cabin)._by_hr_giger_+_Gabriele_Dell'Otto_+_Jeff_Easl.png (768x512, 709.55K)

first

added to my personal spell book

Attached: 1588271788782.png (1024x1024, 1.57M)

Why aren't these ESGRAN models loading?
They are in stable-diffusion-webui-master\ESRGAN, do they need to be in root directory with webui.bat or something?

Attached: file.png (1278x799, 127.76K)

nth for guess the band

Attached: download (47).png (640x448, 512.61K)

they're in the right place but they're not loading because they're the wrong kind of model. RealESRGAN is different from ESRGAN

I saw a video recording of user using it on Linux though, so it's either that one or try looking into koi

boomers would buy calendars of this

hi training guide op here
I recommend using waifu-diff now
I will write the docs soon but if anyone can feel free to do so
If you can name what hardware you have, requirement is now 30GB VRAM
Also would you donate towards training? If so how much and towards what kind of training

>RealESRGAN is different from ESRGAN
computers were a mistake

Using your old Dall-e Mini images as img2img with similar prompts works great to upscale them

the smiths

AAAAAAAA

Attached: 00063-24_k_euler_a_1391214254.png (512x512, 505.27K)

anyone else having trouble with voldy's progressbar and live updates? when i load the ui and run the first generation it works fine, but after if i change some parameters and hit Generate again i just get the bare "loading.." screen

Attached: index6.png (1024x1024, 1.56M)

>30GB VRAM
So close, so close. Someone just needs to shave 6 gb off that. Well probably 9 really. Can that whole memory-swapping thing be used? I know nothing about this but that sequence of words sounds like it would fix things right up.

4090's rumoured to have 48GB isn't it

catch is you probably have to upgrade your power supply too lol

i love horror but this one came out too much for me, yeesh

Attached: Nirvana-1.jpg (1200x800, 153.34K)

eyes up here, user

Attached: messing with ʻʻvanishing_pointʻʻ position in the word order.jpg (512x512, 71.44K)

you deleted your post but as a fellow retard and for my fellow retards, run krita_server.py
also I had to change where it pointed to the config, I don't know fucking why but just go point it at where it actually is, aka right next to the damn .py file

>spellbook
I must already have over 1000 different prompts

retard here, can I alter features based on an image like this? do I just drop the image in img2img and enter a prompt, like smiling man or something?

Attached: 62b46638bedf9aabc6b3c121_Walter white.jpg (600x600, 40.94K)

>KILL THE PLANET,
>KILL THE PLANET,
>KILL THE PLANET

Attached: kplnt.png (1536x1024, 2.67M)

see pic

Attached: 1644786084367.png (512x512, 391.34K)

This is tested on the first version he released, not the newer one as that seems to not work there.

Attached: download (51).png (512x512, 356.72K)

Thank you for leaving your nugget of wisdom with us instead of just bailing

i've been watching this on repeat for an hour. i liked black tape project as a concept, but all the girls were so... ugly. the fact SD doesn't require the actors to look basically the same means deepfake is now obsolete.

Your pic inspired me.

Attached: anime girl soup.png (512x512, 355.64K)

yes, you can also paint a specific area and change only that part with a prompt

your PTO request has been denied, user

Attached: 00036.png (1024x1024, 1.25M)

>eating living anime girl soup
welp, new fetish.

urban exploration photography

Attached: 00454-1274045679_cool_front-lit_sign_I_saw_today_in_chicago,_had_to_have_my_picture_taken!.png (512x512, 408.96K)

my fetish is creating human life with my penis

thanks doc

>DreamStudio Lite
is this the standard DreamStudio or do you not have access to the full one without purchasing credits?

holy kek, i think that one just tied for my favorite with

>fetish
I only make wholesome iyashikei daughterwife anime girl soups, take your sick fantasies elsewhere.

Attached: anime girl soup 2.png (512x512, 379.07K)

Bikini Bottom has seen better days.

Attached: BB.jpg (2560x4224, 2.64M)

Attached: 1644893239435.png (512x512, 467.07K)

Love this user, thank you.

this is gonna be spongebob in 2013

Attached: PRESIDENTS.jpg (4096x1408, 1.18M)

Can anyone fill me in on the lore of this image?

if they aren't for eating why do they look so scared

checkmate atheists

after recommendations in the last thread, i tried separating terms with |

I made it two days ago

imagine her muffled screams as she's being dissolved by stomach acid

Stop it, you are evil.

someone posted a gif of piccolo raping vegeta every day multiple times a day on Any Forums

as such was to happen in the day, instead of ignoring it as a desperate cry for attention, people started calling it a meme until it was hashblocked.

I just installed this, does it only work on the delimited area? lets suppose I want to generate a picture with a raised brow, do I delimit only the brow that's already there, or the whole area where I want to have the raised brow? any other tips?

Attached: Untitled.png (1533x1291, 1014.92K)

Since when did Cannon start making toasters?

Attached: index.jpg (2304x2304, 1.15M)

Attached: download (54).png (640x448, 455.02K)

Anyone got decent Anyas? All mine are horrible deformed

Attached: grid-0158.png (1536x1024, 2.1M)

I know, I was there but it seems like it references some insider joke. Please invite me into the cool boys club, pleaseeee.

Attached: 1412503237072.png (502x499, 250.94K)

Fuck I'm retarded. Running SD on a VM on my server and wanna connect to it from another machine on my LAN. Console window says "To create a public link, set `share=True` in `launch()` but where's launch? I'm running voldy's UI. Already punched a hole in my firewall for the port.

Oh my fucking god I think I see it. Thank you, user-sama

oh you think??