File size: 3,341 Bytes
7df97e8
 
 
 
 
 
9446250
7df97e8
2c06262
76f0746
a533ce1
7df97e8
6028910
 
a533ce1
 
346fc17
 
a533ce1
70a26da
 
 
 
a533ce1
 
 
5edd7d3
 
346fc17
 
 
 
5edd7d3
 
 
 
25602c3
 
5edd7d3
25602c3
 
 
 
 
 
 
 
 
 
 
 
 
 
 
5edd7d3
a533ce1
 
 
7df97e8
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
---
license: mit
---
### I Love Chaos on Stable Diffusion
This is the `<chaos>` concept taught to Stable Diffusion via Textual Inversion. You can load this concept into the [Stable Conceptualizer](https://colab.research.google.com/github/huggingface/notebooks/blob/main/diffusers/stable_conceptualizer_inference.ipynb) notebook. You can also train your own concepts and load them into the concept libraries using [this notebook](https://colab.research.google.com/github/huggingface/notebooks/blob/main/diffusers/sd_textual_inversion_training.ipynb).

This is a concept made from i-love-chaos brand logos, each logo is using a different fractal producing this concept, use it with &lt;chaos&gt;.

Hearts and Butterflies pop up easily with this concept, beside having a very strong red influence on the image.

1000 Steps have been used to train this model.

for supporting us visit [Etsy Shop](https://www.etsy.com/shop/iLoveChaos)

### Usage 

When publishing work, always include the hashtag: #stablediffusion_chaos in any form when publishing to social media/release notes.

####  What does this model do?

For once it makes everything red :). Once in a while it creates heart shapes, this particular model is best used at the end or after the first 5 to 7 keywords, or deantentuate it using (<chaos>) and enclose the keyword in brackets, for attentuating it more, use <chaos>! with the exclamation mark at the end.

It works nicely with 'butterfly', 'heart' and 'tornado'.

In general it gives a nice red-touch to any image but sometimes tries to recreate the text leading to garbage text in the generated image

#### Example:

Prompt Sourced from:

> [Prompt Source 'The 100 Most Beautiful Stable Diffusion Prompts'](https://mpost.io/best-100-stable-diffusion-prompts-the-most-beautiful-ai-text-to-image-prompts/)

Prompt:
  
    new york city, dust storm, cinematic, dramatic, composition,  (<chaos>), sunny sky, brutalist, hyper realistic, epic scale, heart, sense of awe, hypermaximalist, insane level of details, artstation HQ
 
Seed 1
 
<img src="https://huggingface.co/sd-concepts-library/i-love-chaos/resolve/main/2022-10-03-19-35-31-001-1-scale6.50-k_euler_a-001.png" width="512">

And same prompt with non attentuated <chaos>

Seed 1

<img src="https://huggingface.co/sd-concepts-library/i-love-chaos/resolve/main/2022-10-03-19-49-18-001-1-scale6.50-k_euler_a-001.png" width="512">

Seed 3

<img src="https://huggingface.co/sd-concepts-library/i-love-chaos/resolve/main/2022-10-03-19-49-38-003-3-scale6.50-k_euler_a-003.png" width="512">

Seed 363421591

<img src="https://huggingface.co/sd-concepts-library/i-love-chaos/resolve/main/2022-10-03-19-44-56-001-363421591-scale6.50-k_euler_a-001.png" width="512">

 
## Training Data

Here is the new concept you will be able to use as a `style` <chaos> :
![<chaos> 0](https://huggingface.co/sd-concepts-library/i-love-chaos/resolve/main/concept_images/0.jpeg)
![<chaos> 1](https://huggingface.co/sd-concepts-library/i-love-chaos/resolve/main/concept_images/1.jpeg)
![<chaos> 2](https://huggingface.co/sd-concepts-library/i-love-chaos/resolve/main/concept_images/2.jpeg)
![<chaos> 3](https://huggingface.co/sd-concepts-library/i-love-chaos/resolve/main/concept_images/3.jpeg)
![<chaos> 4](https://huggingface.co/sd-concepts-library/i-love-chaos/resolve/main/concept_images/4.jpeg)