EleetModel / README.md
CodeChris's picture
Add Eleet v1.
c855e43
|
raw
history blame
3.59 kB
metadata
license: creativeml-openrail-m

Eleet Model: an anime style stable diffusion mdoel

Eleet model is a block-weighted merged stable diffusion model aiming at generating good quality 2D anime style images that satisfy my personal taste and, hopefully, your taste.

ELEET, Internet slang for the word 'elite'. Its digital transliteration is 31337, which has been the most common ENSD setting in stable diffusion probably since the NAI era. This model is named in honor of this tradition.

Suggested settings

For users who know little about stable diffusion settings, I recommend:

  • Prompt: Start with masterpiece, best quality. Personally I also like to add:

    • high-resolution or aesthetic
    • Photography pharses such as cinematic lighting, sharp focus, etc.
  • Negative prompt:

    (worst quality, low quality:1.4), lowres, bad anatomy, (child, loli), (blurry, depth of field, bokeh), (text, logo, watermark, signature, username)

  • Sampler: DPM++ 2M Karras

  • CFG Scale: 7~12

  • Steps: 20

  • Clip skip: 2

  • No external VAE needed

Samples

Due to my GPU limitation, I couldn't generate ultra-high resolution images and I didn't use Highres Fix very often.

Sample 1: Scenery. txt2img+highres, (640x384) x1.5.

masterpiece, best quality, aesthetic, highres RAW photo, landscape photography, wide shot, from below, scenery, sunrise, blue sky, clouds, lake, reflection, sun, trees, floating leaves, ripples, foreground interest, depth of field, cinematic lighting, asymmetric composition, professional shadows, sharp focus, lens flare

Sample 2: Anime girl. txt2img, 576x832.

masterpiece, best quality, 1girl, solo, angry, brown eyes, blue hair, bangs, sidelocks, half updo, clenched hand, parted lips, [bodysuit|armored dress], long sleeves, gloves, armor, looking to the side, clenched teeth, (cowboy shot), night, full moon, forest, underexposure, professional lighting

Click here to expand and see more samples

Sample 3: Scenery. txt2img, 832x576.

masterpiece, best quality, aesthetic, 1girl, :d, purple eyes, fox ears, gold hair, two side up, crop top, off-shoulder jacket, green pleated skirt, thighhighs, garter belt, skindentation, breasts, bare shoulders, midriff, (cowboy shot), looking at viewer, waterfall, mountains

Sample 4: Anime girl. txt2img+highres, (384x640) x1.5.

masterpiece, best quality, aesthetic, highres RAW photo, cool color tone, wide shot, scenery, snow, blue sky, cityscape, skyline, buildings, rooftop, roads, cinematic lighting, professional shadows, sharp focus

Merge ideas

The weights for merging Eleet model were optimized through an auotmatic processure with scoring, but I didnt' necessarily pick the best-scored one as the final version. Instead, I will evaluate a number of high-scored candidates and score their outputs manually by myself. I conducted the evaluation mostly on anime girls topics (no doubt) but also considered the model performance on scenary images.

Following models are used in the Eleet merge:

License

CreativeML OpenRAIL-M.