EvkoGS commited on
Commit
fcc714c
1 Parent(s): 8a13ba5

adding images and updating readme

Browse files
README.md CHANGED
@@ -1,4 +1,129 @@
1
  ---
2
- library_name: diffusers
3
  license: openrail++
 
 
 
 
 
 
 
 
4
  ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
  ---
 
2
  license: openrail++
3
+ language:
4
+ - en
5
+ tags:
6
+ - stable-diffusion-xl
7
+ - text-to-image
8
+ - art
9
+ - fine-art
10
+ inference: true
11
  ---
12
+
13
+ <style>
14
+ @import url('https://fonts.googleapis.com/css2?family=Lora&display=swap');
15
+
16
+ .title-container {
17
+ display: flex;
18
+ justify-content: center;
19
+ align-items: center;
20
+ height: 20vh;
21
+ background-color: #000;
22
+ }
23
+
24
+ .title {
25
+ font-family: 'Lora', sans-serif;
26
+ font-size: 4em;
27
+ text-align: center;
28
+ color: #fff;
29
+ text-transform: uppercase;
30
+ letter-spacing: 0.1em;
31
+ padding: 0.5em 0;
32
+ }
33
+
34
+ .image-grid {
35
+ display: grid;
36
+ grid-template-columns: 1fr 1fr;
37
+ grid-template-rows: 1fr 1fr auto;
38
+ gap: 0; /* Removed gap to eliminate space between images */
39
+ background-color: #000; /* Changed to black background */
40
+ }
41
+
42
+ .image-grid img {
43
+ width: 100%;
44
+ height: 100%;
45
+ object-fit: cover;
46
+ display: block; /* Ensures no extra space around images */
47
+ }
48
+
49
+ .image-grid .wide-image {
50
+ grid-column: 1 / -1;
51
+ }
52
+ </style>
53
+
54
+ <div class="title-container">
55
+ <h1 class="title">Artisan XL</h1>
56
+ </div>
57
+
58
+ <div class="image-grid">
59
+ <img src="images/Picasso.webp" alt="Square image 1">
60
+ <img src="images/Repin.webp" alt="Square image 2">
61
+ <img src="images/Van Gogh.webp" alt="Square image 3">
62
+ <img src="images/Vermeer.webp" alt="Square image 4">
63
+ <img src="images/long_van_gogh.webp" alt="Wide image" class="wide-image">
64
+ </div>
65
+
66
+
67
+ # Artisan XL: Unparalleled Fine Art Generation
68
+
69
+ Artisan XL is a state-of-the-art text-to-image model that pushes the boundaries of AI-generated fine art. Trained on an extensive dataset of 200,000 high-resolution oil paintings, this model offers unparalleled style fidelity for a wide range of artistic styles throughout human history.
70
+
71
+ ## Key Features
72
+
73
+ - Trained for 1200 A100 hours on 8 A100 GPUs
74
+ - Captioned with Claude Sonnet 3.5 for precise style understanding
75
+ - Exceptional style fidelity for artists like Picasso, Van Gogh, Rembrandt, Klimt, Repin, Serov, Kandinsky, and many more
76
+ - Supports all basic SDXL tools and community techniques
77
+
78
+ ## Usage Tips
79
+
80
+ 1. **Prompt Engineering**: We recommend using long, detailed prompts with weighting. Ideally, generate prompts using Claude Sonnet or Haiku for best results.
81
+ 2. **Negative Prompts**: Use empty or very short negative prompts. Common negative keywords include: `simplified, rough, blurred, crude, imperfections, sketch`
82
+ 3. **Settings**:
83
+ - Works well with guidance limiter
84
+ - Recommended CFG scale: 3.0-5.5
85
+
86
+ ## Model Variants
87
+
88
+ 1. **Full Model**: Complete SDXL fine-tuned model
89
+ 2. **LoRA Variant**:
90
+ - Linear and Conv modules
91
+ - Rank 128
92
+ - Provides 95% of full model capabilities
93
+
94
+ ### Multi Aspect Resolution
95
+
96
+ This model supports generating images at the following dimensions:
97
+
98
+ | Dimensions | Aspect Ratio |
99
+ |-------------------|-----------------|
100
+ | `1024 x 1024` | 1:1 Square |
101
+ | `1152 x 896` | 9:7 |
102
+ | `896 x 1152` | 7:9 |
103
+ | `1216 x 832` | 19:13 |
104
+ | `832 x 1216` | 13:19 |
105
+ | `1344 x 768` | 7:4 Horizontal |
106
+ | `768 x 1344` | 4:7 Vertical |
107
+ | `1536 x 640` | 12:5 Horizontal |
108
+ | `640 x 1536` | 5:12 Vertical |
109
+
110
+ ### Acknowledgements
111
+
112
+ The development and release of Artisan XL would not have been possible without the invaluable contributions and support from the following individuals and organizations:
113
+
114
+ - **[Kohya SS](https://github.com/kohya-ss)**: For training scripts.
115
+
116
+ ## Limitations
117
+
118
+ While Artisan XL offers exceptional capabilities in fine art generation, it's important to note:
119
+
120
+ 1. The model specializes in fine art styles and may not perform as well for other genres.
121
+ 2. Results can vary based on prompt quality and settings.
122
+ 3. As with all AI models, occasional unexpected outputs may occur.
123
+
124
+ ## Contact
125
+
126
+ For any inquiries, collaborations, or custom training requests, please contact:
127
+ grigory@theartisan.ai
128
+
129
+ We're excited to see what you create with Artisan XL! Happy art-making!
images/Picasso.webp ADDED
images/Repin.webp ADDED
images/Van Gogh.webp ADDED
images/Vermeer.webp ADDED
images/long_van_gogh.webp ADDED