部分模型官方用法文档

自用，备忘

7th Layer

(Important Notice:1.6)
default CFG Scale : 7 ± 5
default Sampler : DPM++ 2M Karras
default Steps : 25
Negative prompt : (worst quality:1.4), (low quality:1.4) , (monochrome:1.1),
Don’t write a lot of “Negative prompt”.
e.g.
4IYqlYq

gsdf/Counterfeit-V2.0

e.g.
((masterpiece, best quality)), a girl, flower, dress, solo, lying, rain, butterfly, bug, water, bangs, frills, breasts, long hair, white dress, short sleeves, hair ornament, on back, outstretched arm, frilled dress, arm up, white flower, hair flower, grey eyes, white hair,looking away,
Negative prompt: (low quality, worst quality:1.4), (bad anatomy), (inaccurate limb:1.2),bad composition, inaccurate eyes, extra digit,fewer digits,(extra arms:1.2),
Steps: 20, Sampler: DPM++ SDE Karras, CFG scale: 8, Size: 640x384, Denoising strength: 0.6, Clip skip: 2, Hires upscale: 2, Hires upscaler: Latent

((masterpiece, best quality)),a girl, solo, hat, blush,long hair, skirt, beret, sitting, bangs, socks, wariza, pink hair, light blue eyes, black headwear,holding,rifle,weapon, looking at viewer, white sailor collar, school uniform, closed mouth, black hat, sailor collar, holding weapon, long sleeves, pleated skirt, white socks,indoors,industrial
Negative prompt: (low quality, worst quality:1.4), (bad anatomy), (inaccurate limb:1.2),bad composition, inaccurate eyes, extra digit,fewer digits,(extra arms:1.2),
Steps: 20, Sampler: DPM++ SDE Karras, CFG scale: 8, Size: 576x384 or 576x448, Denoising strength: 0.6, Clip skip: 2, Hires upscale: 2, Hires upscaler: Latent

(masterpiece, best quality)),a girl, solo, twintails, shirt, skirt, petals, bowtie, earrings, jewelry, bangs, black hair, hair ornament, hair ribbon, red ribbon, red eyes, long hair, open mouth, white shirt, multicolored hair, black skirt, red hair, long sleeves, pink bowtie, hair between eyes, looking at viewer, collared shirt, upper body, hand up, falling petals, depth of field, strong bloom, red background
Negative prompt: (low quality, worst quality:1.4), (bad anatomy), (inaccurate limb:1.2),bad composition, inaccurate eyes, extra digit,fewer digits,(extra arms:1.2),
Steps: 20, Sampler: DPM++ SDE Karras, CFG scale: 8, Size: 640x384, Denoising strength: 0.6, Clip skip: 2, Hires upscale: 2, Hires upscaler: Latent

nuigurumi/basil_mix

Model Description

merged model.
realistic texture and Asian face.
designed to maintain a responsive reaction to danbooru based prompts.

Recommendations

VAE: “vae-ft-mse-840000” from StabilityAI
Prompting: Simple prompts are better. Large amounts of quality tags and negative prompts can have negative effects.

Linaqruf/anything-v3.0v

Anything V3
Welcome to Anything V3 - a latent diffusion model for weebs. This model is intended to produce high-quality, highly detailed anime style with just a few prompts. Like other anime-style Stable Diffusion models, it also supports danbooru tags to generate images.

e.g. 1girl, white hair, golden eyes, beautiful eyes, detail, flower meadow, cumulonimbus clouds, lighting, detailed sky, garden

Anime Girl:

1girl, brown hair, green eyes, colorful, autumn, cumulonimbus clouds, lighting, blue sky, falling leaves, garden
Steps: 50, Sampler: DDIM, CFG scale: 12

Scenery:

scenery, shibuya tokyo, post-apocalypse, ruins, rust, sky, skyscraper, abandoned, blue sky, broken window, building, cloud, crane machine, outdoors, overgrown, pillar, sunset
Steps: 50, Sampler: DDIM, CFG scale: 12

nitrosocke/elden-ring-diffusion

Elden Ring Diffusion
This is the fine-tuned Stable Diffusion model trained on the game art from Elden Ring. Use the tokens elden ring style in your prompts for the effect.

Prompt and settings for portraits:

elden ring style portrait of a beautiful woman highly detailed 8k elden ring
style Steps: 35, Sampler: DDIM, CFG scale: 7, Seed: 3289503259, Size: 512x704

Prompt and settings for landscapes:

elden ring style dark blue night (castle) on a cliff dark night (giant birds) elden ring style Negative prompt: bright day
Steps: 30, Sampler: DDIM, CFG scale: 7, Seed: 350813576, Size: 1024x576

WarriorMama777/OrangeMixs

OrangeMixs
“OrangeMixs” shares various Merge models that can be used with StableDiffusionWebui:Automatic1111 and others.

How to use

VAE: orangemix.vae.pt
~~Prompts can be long or short~~
As simple as possible is good. Do not add excessive detail prompts. Start with just this negative propmt.
(worst quality, low quality:1.4)
Sampler: “DPM++ SDE Karras” is good
Steps: forTest: 12～ ,illustration: 20～
Clipskip: 1 or 2
Upscaler : Latenet (nearest-exact)
CFG Scale : 5 or 6 (4～8)
Denoise strength: 0.5 (0.45~0.6)
If you use 0.7～, the picture will change too much.
If below 0.45, Block noise occurs.

How to choice models

_sfw : SFW😉
_nsfw : SFW ～ Soft NSFW🥰
_hard : SFW ～ hard NSFW👄

e.g.

prompt:
masterpiece, best quality, ultra-detailed, illustration, 2girls, sitting, playful, gaming, messy room, teenage, 15 years old, lighthearted, cozy, relaxed, cheerful, fondness, love, friendship, cute, touching, computer game, controllers, smiles, joy, laughter, comfortable, slippers, pajamas, messy hair, tousled, disarray, cluttered, toys, posters, pillows, blankets, lamp, desk, chair, cozy atmosphere, warm lighting, bright colors, soft pastels, flowers, plants, books, headphones, snacks, soda, energy drinks, manga, novels, plushies, figurines, posters, pictures, posters, wall scrolls, stickers, decorations, bed, blankets, pillows, stuffed animals, cozy blankets, warm blankets, comfortable clothes, casual attire, leisure wear, sweatshirt, sweatpants, shorts, t-shirt, tank top, socks,
negative prompt:
nsfw, (worst quality, low quality, extra digits:1.4)),
Image size 512x768
Steps 20
Sampler DPM++ SDE Karras
CFG scale 8
Seed2745740002
Model AbyssOrangeMix2_sfw
Denoising strength 0.5
ENSD 31337
Hires upscale 1.8
Hires steps 20
Hires upscaler Latent (nearest-exact)

wlop

This model works better with the anything.vae to generate better coloring

Putting ENSD(eta noise seed delta) at 31337 and clip skip at 2 or 3 in settings is also recommended

A low CFG between 5 and 7 usually gives better results

Thumbnail image settings
Steps: 20, Sampler: DPM++ 2S a Karras, CFG scale: 7, Seed: 3455876617, Size: 776x648, Model hash: 8c9e045b, Denoising strength: 0.6, Clip skip: 2, ENSD: 31337, First pass size: 0x0

Upscale: 4, visibility: 1.0, model:R-ESRGAN 4x+ Anime6B
e.g.

Prompt:
8k,(sexy:1.2), wizard,lingerie, upper body portrait of a girl , hair tied in bun, particle effects medium breasts, detailed shop interior, vibrant jars , Albert Lynch, J. C. Leyendecker, Ruan Jia, Gaston Bussiere, Alexandre Cabanel, WLOP, Masterpiece, volumetric lighting, posing,hd, (chromatic aberration), detailed pupils, dual tone, fov
Negative prompt:
obese, (ugly:1.3), (duplicate:1.3), (morbid), (mutilated), out of frame, extra fingers, mutated hands, (poorly drawn hands), (poorly drawn face), (mutation:1.3), (deformed:1.3), (amputee:1.3), blurry, bad anatomy, bad proportions, (extra limbs), cloned face, (disfigured:1.3), gross proportions, (malformed limbs), (missing arms), (missing legs), (extra arms), (extra legs), mutated hands, (fused fingers), (too many fingers), (long neck:1.3), lowres, text, error, cropped, worst quality, low quality, normal quality, jpeg artifacts, signature, watermark, username, blurry, black and white, monochrome, censored,empty
Sampler: DPM++ 2S a Karras
CFG scale 7
Steps 20
Seed 3455876617

Himitsu - HaneRu V1.0

模型介绍

Base：AnythingV3 - fp16
Dataset：Himitsu，HaneRu(About 600 pictures)
Training method：Finetuning
e.g.

Prompt:
1girl,Mi-Ha5-7500,arms_behind_back, bangs, bare_shoulders, bdsm, black_gloves, black_legwear, blindfold, blush, bondage, bound, bound_ankles, bound_arms, bound_legs, bound_wrists, box_tie, breast_bondage, breasts, chained, chair, crotch_rope, crow, dove, feathers, frogtie, gloves, handcuffs, looking_at_viewer, nose_blush, open_mouth, outdoors, panties, red_rope, restrained, ribbon_bondage, rope, shibari, shibari_over_clothes, sidelocks, skindentation, suspension, tape, tickling, tied_up_(nonsexual), torn_clothes, twintails, underwear
Negative prompt:
tape_gag,gagged,gag,cleave_gag,ring_gag,lowres, bad_anatomy, bad_hands, text, error, missing_fingers, extra_digit, fewer_digits, cropped, worst_quality, low_quality, normal_quality, jpeg_artifacts, signature, watermark, username, blurry
CFG scale 12
Steps 20
Sampler: Euler a
Seed 3219661566

Grapefruit (hentai model)

grapefruit (general hentai model)
grape (gape and insertion model -> old and only for merging good, has big flaws in compare to grapefruit)
Grapefruit aims to be a hentai model with a bright and more „softer“ artstyle. Compares with other mixes: Simple dildo vaginal, the same now with dildo anal and a down below close up. Model is also good for SFW images.
Use a vae with it. AnythingV3 vae as extra download here, down arrow next to the download button. But you can use any vae you like. Savetensor and the vae file must be named the same -> MODEL.savetensor and MODEL.vae.pt
e.g.

Prompt:
masterpiece, best quality, 1girl, pink eyes, long hair, black hair, (grapefruit), (temple in background), sitting, kimono, medium breasts, topless, light smile, arms behind back, [nipples],
Negative prompt:
(worst quality, low quality:1.4), logo, text, monochrome,
ENSD: 31337
Size: 392x512
Seed: 4272347218
Steps: 20
Sampler: DDIM
CFG scale: 8
Clip skip: 2
Hires upscale: 2.35
Hires upscaler: Latent (nearest-exact)
Denoising strength: 0.77