Update README.md
Browse files
README.md
CHANGED
@@ -20,7 +20,6 @@ widget:
|
|
20 |
negative_prompt: nsfw, lowres, bad anatomy, bad hands, text, error, missing fingers, extra digit, fewer digits, cropped, worst quality, low quality, normal quality, jpeg artifacts, signature, watermark, username, blurry, artist name
|
21 |
example_title: 1boy
|
22 |
---
|
23 |
-
|
24 |
<style>
|
25 |
.title-container {
|
26 |
display: flex;
|
@@ -40,19 +39,23 @@ widget:
|
|
40 |
background: transparent;
|
41 |
}
|
42 |
|
|
|
|
|
|
|
|
|
|
|
43 |
.title span {
|
44 |
background: -webkit-linear-gradient(45deg, #7ed56f, #28b485);
|
45 |
-webkit-background-clip: text;
|
46 |
-webkit-text-fill-color: transparent;
|
47 |
}
|
48 |
-
|
49 |
.subtitle {
|
50 |
font-size: 1.5em;
|
51 |
text-align: center;
|
52 |
color: #777;
|
53 |
font-family: 'Helvetica Neue', sans-serif;
|
54 |
text-transform: uppercase;
|
55 |
-
margin-top:
|
56 |
letter-spacing: 0.2em;
|
57 |
background: transparent;
|
58 |
}
|
@@ -62,7 +65,6 @@ widget:
|
|
62 |
-webkit-background-clip: text;
|
63 |
-webkit-text-fill-color: transparent;
|
64 |
}
|
65 |
-
|
66 |
.custom-table {
|
67 |
table-layout: fixed;
|
68 |
width: 100%;
|
@@ -76,22 +78,19 @@ widget:
|
|
76 |
padding: 10px;
|
77 |
box-shadow: 0px 0px 0px 0px rgba(0, 0, 0, 0.15);
|
78 |
}
|
79 |
-
|
80 |
.custom-image-container {
|
81 |
position: relative;
|
82 |
width: 100%;
|
83 |
margin-bottom: 0em;
|
84 |
overflow: hidden;
|
85 |
border-radius: 10px;
|
86 |
-
transition: transform .
|
87 |
/* Smooth transition for the container */
|
88 |
}
|
89 |
-
|
90 |
.custom-image-container:hover {
|
91 |
-
transform: scale(1.
|
92 |
/* Scale the container on hover */
|
93 |
}
|
94 |
-
|
95 |
.custom-image {
|
96 |
width: 100%;
|
97 |
height: auto;
|
@@ -100,12 +99,10 @@ widget:
|
|
100 |
transition: transform .7s;
|
101 |
margin-bottom: 0em;
|
102 |
}
|
103 |
-
|
104 |
.nsfw-filter {
|
105 |
filter: blur(8px); /* Apply a blur effect */
|
106 |
transition: filter 0.3s ease; /* Smooth transition for the blur effect */
|
107 |
}
|
108 |
-
|
109 |
.custom-image-container:hover .nsfw-filter {
|
110 |
filter: none; /* Remove the blur effect on hover */
|
111 |
}
|
@@ -141,7 +138,7 @@ widget:
|
|
141 |
/* Fallback for browsers that do not support this effect */
|
142 |
text-shadow: 2px 2px 4px rgba(0, 0, 0, 0.7);
|
143 |
/* Enhanced text shadow for better legibility */
|
144 |
-
|
145 |
.overlay-subtext {
|
146 |
font-size: 0.75em;
|
147 |
margin-top: 0.5em;
|
@@ -151,10 +148,8 @@ widget:
|
|
151 |
.overlay,
|
152 |
.overlay-subtext {
|
153 |
text-shadow: 2px 2px 4px rgba(0, 0, 0, 0.5);
|
154 |
-
}
|
155 |
-
|
156 |
</style>
|
157 |
-
|
158 |
<h1 class="title">
|
159 |
<span>Animagine XL 3.1</span>
|
160 |
</h1>
|
@@ -188,7 +183,6 @@ widget:
|
|
188 |
</td>
|
189 |
</tr>
|
190 |
</table>
|
191 |
-
|
192 |
**Animagine XL 3.1** is the latest version of the sophisticated open-source anime text-to-image model, building upon the capabilities of its predecessor, Animagine XL 3.0. Developed based on Stable Diffusion XL, this iteration boasts superior image generation with notable improvements in hand anatomy, efficient tag ordering, and enhanced knowledge about anime concepts. Unlike the previous iteration, we focused to make the model learn concepts rather than aesthetic.
|
193 |
## What’s New in Animagine XL 3.1 ?
|
194 |
## Aesthetic Tags
|
@@ -235,13 +229,11 @@ from diffusers import (
|
|
235 |
EulerAncestralDiscreteScheduler,
|
236 |
AutoencoderKL
|
237 |
)
|
238 |
-
|
239 |
# Load VAE component
|
240 |
vae = AutoencoderKL.from_pretrained(
|
241 |
"madebyollin/sdxl-vae-fp16-fix",
|
242 |
torch_dtype=torch.float16
|
243 |
)
|
244 |
-
|
245 |
# Configure the pipeline
|
246 |
pipe = StableDiffusionXLPipeline.from_pretrained(
|
247 |
"cagliostrolab/animagine-xl-3.1",
|
@@ -251,11 +243,9 @@ pipe = StableDiffusionXLPipeline.from_pretrained(
|
|
251 |
)
|
252 |
pipe.scheduler = EulerAncestralDiscreteScheduler.from_config(pipe.scheduler.config)
|
253 |
pipe.to('cuda')
|
254 |
-
|
255 |
# Define prompts and generate image
|
256 |
prompt = "1girl, arima kana, oshi no ko, solo, upper body, v, smile, looking at viewer, outdoors, night"
|
257 |
negative_prompt = "nsfw, lowres, bad anatomy, bad hands, text, error, missing fingers, extra digit, fewer digits, cropped, worst quality, low quality, normal quality, jpeg artifacts, signature, watermark, username, blurry, artist name"
|
258 |
-
|
259 |
image = pipe(
|
260 |
prompt,
|
261 |
negative_prompt=negative_prompt,
|
@@ -433,4 +423,4 @@ Based on Animagine XL 3.0, Animagine XL 3.1 falls under [Fair AI Public License
|
|
433 |
3. **Distribution Terms:** Any distribution must be under this license or another with similar rules.
|
434 |
4. **Compliance:** Non-compliance must be fixed within 30 days to avoid license termination, emphasizing transparency and adherence to open-source values.
|
435 |
|
436 |
-
The choice of this license aims to keep Animagine XL 3.1 open and modifiable, aligning with open source community spirit. It protects contributors and users, encouraging a collaborative, ethical open-source community. This ensures the model not only benefits from communal input but also respects open-source development freedoms.
|
|
|
20 |
negative_prompt: nsfw, lowres, bad anatomy, bad hands, text, error, missing fingers, extra digit, fewer digits, cropped, worst quality, low quality, normal quality, jpeg artifacts, signature, watermark, username, blurry, artist name
|
21 |
example_title: 1boy
|
22 |
---
|
|
|
23 |
<style>
|
24 |
.title-container {
|
25 |
display: flex;
|
|
|
39 |
background: transparent;
|
40 |
}
|
41 |
|
42 |
+
h1.title {
|
43 |
+
margin-bottom: 0px;
|
44 |
+
line-height: 0.4em;
|
45 |
+
}
|
46 |
+
|
47 |
.title span {
|
48 |
background: -webkit-linear-gradient(45deg, #7ed56f, #28b485);
|
49 |
-webkit-background-clip: text;
|
50 |
-webkit-text-fill-color: transparent;
|
51 |
}
|
|
|
52 |
.subtitle {
|
53 |
font-size: 1.5em;
|
54 |
text-align: center;
|
55 |
color: #777;
|
56 |
font-family: 'Helvetica Neue', sans-serif;
|
57 |
text-transform: uppercase;
|
58 |
+
margin-top: 0em;
|
59 |
letter-spacing: 0.2em;
|
60 |
background: transparent;
|
61 |
}
|
|
|
65 |
-webkit-background-clip: text;
|
66 |
-webkit-text-fill-color: transparent;
|
67 |
}
|
|
|
68 |
.custom-table {
|
69 |
table-layout: fixed;
|
70 |
width: 100%;
|
|
|
78 |
padding: 10px;
|
79 |
box-shadow: 0px 0px 0px 0px rgba(0, 0, 0, 0.15);
|
80 |
}
|
|
|
81 |
.custom-image-container {
|
82 |
position: relative;
|
83 |
width: 100%;
|
84 |
margin-bottom: 0em;
|
85 |
overflow: hidden;
|
86 |
border-radius: 10px;
|
87 |
+
transition: transform .4s;
|
88 |
/* Smooth transition for the container */
|
89 |
}
|
|
|
90 |
.custom-image-container:hover {
|
91 |
+
transform: scale(1.17);
|
92 |
/* Scale the container on hover */
|
93 |
}
|
|
|
94 |
.custom-image {
|
95 |
width: 100%;
|
96 |
height: auto;
|
|
|
99 |
transition: transform .7s;
|
100 |
margin-bottom: 0em;
|
101 |
}
|
|
|
102 |
.nsfw-filter {
|
103 |
filter: blur(8px); /* Apply a blur effect */
|
104 |
transition: filter 0.3s ease; /* Smooth transition for the blur effect */
|
105 |
}
|
|
|
106 |
.custom-image-container:hover .nsfw-filter {
|
107 |
filter: none; /* Remove the blur effect on hover */
|
108 |
}
|
|
|
138 |
/* Fallback for browsers that do not support this effect */
|
139 |
text-shadow: 2px 2px 4px rgba(0, 0, 0, 0.7);
|
140 |
/* Enhanced text shadow for better legibility */
|
141 |
+
}
|
142 |
.overlay-subtext {
|
143 |
font-size: 0.75em;
|
144 |
margin-top: 0.5em;
|
|
|
148 |
.overlay,
|
149 |
.overlay-subtext {
|
150 |
text-shadow: 2px 2px 4px rgba(0, 0, 0, 0.5);
|
151 |
+
}
|
|
|
152 |
</style>
|
|
|
153 |
<h1 class="title">
|
154 |
<span>Animagine XL 3.1</span>
|
155 |
</h1>
|
|
|
183 |
</td>
|
184 |
</tr>
|
185 |
</table>
|
|
|
186 |
**Animagine XL 3.1** is the latest version of the sophisticated open-source anime text-to-image model, building upon the capabilities of its predecessor, Animagine XL 3.0. Developed based on Stable Diffusion XL, this iteration boasts superior image generation with notable improvements in hand anatomy, efficient tag ordering, and enhanced knowledge about anime concepts. Unlike the previous iteration, we focused to make the model learn concepts rather than aesthetic.
|
187 |
## What’s New in Animagine XL 3.1 ?
|
188 |
## Aesthetic Tags
|
|
|
229 |
EulerAncestralDiscreteScheduler,
|
230 |
AutoencoderKL
|
231 |
)
|
|
|
232 |
# Load VAE component
|
233 |
vae = AutoencoderKL.from_pretrained(
|
234 |
"madebyollin/sdxl-vae-fp16-fix",
|
235 |
torch_dtype=torch.float16
|
236 |
)
|
|
|
237 |
# Configure the pipeline
|
238 |
pipe = StableDiffusionXLPipeline.from_pretrained(
|
239 |
"cagliostrolab/animagine-xl-3.1",
|
|
|
243 |
)
|
244 |
pipe.scheduler = EulerAncestralDiscreteScheduler.from_config(pipe.scheduler.config)
|
245 |
pipe.to('cuda')
|
|
|
246 |
# Define prompts and generate image
|
247 |
prompt = "1girl, arima kana, oshi no ko, solo, upper body, v, smile, looking at viewer, outdoors, night"
|
248 |
negative_prompt = "nsfw, lowres, bad anatomy, bad hands, text, error, missing fingers, extra digit, fewer digits, cropped, worst quality, low quality, normal quality, jpeg artifacts, signature, watermark, username, blurry, artist name"
|
|
|
249 |
image = pipe(
|
250 |
prompt,
|
251 |
negative_prompt=negative_prompt,
|
|
|
423 |
3. **Distribution Terms:** Any distribution must be under this license or another with similar rules.
|
424 |
4. **Compliance:** Non-compliance must be fixed within 30 days to avoid license termination, emphasizing transparency and adherence to open-source values.
|
425 |
|
426 |
+
The choice of this license aims to keep Animagine XL 3.1 open and modifiable, aligning with open source community spirit. It protects contributors and users, encouraging a collaborative, ethical open-source community. This ensures the model not only benefits from communal input but also respects open-source development freedoms.
|