Chaosmix

Prompt

official art, unity 8k wallpaper, ultra detailed, beautiful and aesthetic, High quality, beautiful, masterpiece, best quality,

Negative Prompt

easynegative, (worst quality, low quality, normal quality:2),lowres, normal quality, split_screen, framed border, childlike paintings, bad lighting, sketches, (monochrome), (grayscale), out of frame

Prompt

1 boy,solo,action-packed,<lora:chaosmix_0.1>,chaosmix,chaos,horror,Alphonse Mucha,blood,long sword,

Negative Prompt

(worst quality, low quality, extra digits:1.2),

Prompt

1girl, solo, action-packed, gothic, <lora:chaosmix_0.1-000002:1>, chaosmix, chaos, horror

Negative Prompt

ngng00, FastNegativeV2

Prompt

1man, solo, sunset, <lora:chaosmix_0.1-000002:0.8>, chaosmix, chaos, horror,

Negative Prompt

(worst quality, low quality:1.4), By bad artist -neg, (badhandv4:1.4), FastNegativeV2, EasyNegative, (bad-hands-5:1.4), rmadanegative402_sd15-neg, realisticvision-negative-embedding, (monochrome), watermark, username, patreon username, patreon logo,

Prompt

1man, solo, convention, <lora:chaosmix_0.1-000002:0.8>, chaosmix, chaos, horror

Negative Prompt

(worst quality, low quality, extra digits:1.2)

Prompt

1girl, solo, bookstore, <lora:chaosmix_0.1-000002:1>, chaosmix, chaos, horror

Negative Prompt

(worst quality, low quality, extra digits:1.2)

Model description

Version 0.1

Lora adds a bit of "chaos" to your art

weight: 0.8 to 1

tag: chaosmix

you can also add "horror", "chaos" and other similar tags for effect

I trained several epochs, but the 2nd one seemed to work better, gives fewer artifacts, I searched for the right weights and epochs for a long time, found more or less optimal, but you can still search

The 3rd epoch feels very overtrained, but perhaps after playing with the scales it will be possible to achieve better results than on the 2nd epoch

perhaps a little later I will adjust the dataset and training settings of Laura and perezal

from https://civitai.com/models/98602/chaosmix

Trigger words

You should use chaosmix to trigger the image generation.

Download model

Weights for this model are available in Safetensors format.

Download them in the Files & versions tab.