cooking repertoire

Table of Contents

1. WHAT IS THIS

  • a meta index, an index of intentions
  • an attempt at avoiding research duplication
  • indexes and tag-searchs are useful, but often what one needs are recipe-like tips, and most likely something done by someone else before
    • besides, many lora-indexes so far are guitly of not having visual examples

2. WHAT WOULD YOU DO

  • add your own recipe using your intent as heading

3. INGREDIENTS

  • rule of thumb use clipskip1 for “realistic” and clipskip2 for anime/cartoon

3.2. LORAS ORGANIZED BY INTENT

3.2.1. ACTIVITY

3.2.2. HIGH DEFINITION

3.2.4. SIMPLE STYLE

3.2.6. FACELESS

3.2.7. TRANSPARENT

3.2.7.1. SEE-THROUGH CLOTHES
3.2.7.2. TRANSPARENT PLASTIC

I tried see-through plastic, translucent, transparent. hologram and clear plastic “transparent PVC” and “transparent vinyl”

3.2.9. BLACK AND WHITE MANGA, LINEART

https://civitai.com/tag/lineart https://civitai.com/tag/flat%20color

3.2.9.1. INVERTED INMERSION

3.2.13. PIXEL

https://civitai.com/models/183354/sdxl-ms-paint-portraits https://civitai.com/models/105025/hyper-light-drifter-lora https://civitai.com/models/129879/dreamshaper-pixelart

kezime is a pixelart artist so the kezime lora makes the pixels (pixel-anime merge) {{artistnamako (takorin)(…)

3.3. PROMPTS

3.3.1. SIMPLE OIL

Armand Guillaumin

3.3.2. OBJECTS

  • dragons: RaynaenvisionXL.safetensorspainting by ashley wood00009.png

3.3.3. DA-VINCI LIKE

Medical illustration, biological illustration Max Brödel Andreas Vesalius Illustrated textbook victorian illustrators Printmaking Vintage Old anatomy drawings Christopher Wren artwork

Robert Hooke, Micrographia da-vinci style black ink on paper lithography

scientific journal of new species anatomy, fantasy, detailed notes, anatomical drawing of monsters https://civitai.com/models/199836/davinci-sketch

3.3.4. GURO ANATOMY

  • ecorche(escorche) style = flayed style

Apterus Illustrations zdzisław beksiński scorn

3.3.5. TILED-CITY GENERATIONS

  • https://twitter.com/DiscoverStabDif
    • My pipeline process is generate a bunch of tiles based on some prompt. Stitch these tiles together and then do an image-to-image using the same prompt with like a 0.5 strength to preserve some of the initial tile info while fusing the tiles together to create a coherent scene. I then do a series of image-to-image generations while randomly moving 10 random pixels around the base image with each frame - this changes the output image slightly. Continuing this forward and you get this cool scene transformation. More can be found here

3.3.6. ICE SCULPTURES

3.3.6.1. ONE

DPM++2SaKarras CFG5 Clipskip2 Sample33

medium shot of an enormous extremely detailed crystalclear ice-sculpted mythicalcreature, early morning, soft lighting, backlit, cinematic photo, 35mm f/11 photograph, film, professional, 4k,

(worst quality, bad quality, normal quality :1.5), drawing, painting, crayon, sketch, graphite, impressionist, noisy, blurry, soft, deformed,

3.3.6.2. TWO

medium shot of an enormous extremely detailed crystalclear ice-sculpted mythicalcreature, early morning, soft lighting, backlit, cinematic photo, 35mm f/11 photograph, film, professional, 4k,

Steps: 30, Sampler: DPM++ 2M SDE Karras, CFG scale: 7, Seed: 3494717827.0, Size: 1152x896, Model: juggernautXLversion2113240.safetensors, Clip skip: 3, Version: v1.5.1

3.3.7. COLORFUL LANDSCAPE WITCH COLORS

vaporwave Alena Aenami

3.3.8. DOOM HERCULES

Frank frazetta

Author: Tekakutli

Created: 2024-03-04 Mon 08:55