File size: 6,869 Bytes
380dc05
 
292dafa
 
 
 
 
380dc05
292dafa
 
 
 
 
 
 
 
 
2b0b9b5
292dafa
 
 
f8a5ff6
 
61d32b3
c5205c3
f8a5ff6
 
 
 
 
 
 
 
61d32b3
f8a5ff6
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
61d32b3
4b8db57
c5205c3
17f134e
 
 
 
 
2b0b9b5
 
 
 
 
 
 
17f134e
 
2b0b9b5
 
 
 
 
 
 
 
 
 
17f134e
61d32b3
4b8db57
c5205c3
17f134e
c030390
 
61d32b3
c030390
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
f8a5ff6
c030390
 
 
 
 
9b353f5
c030390
 
 
292dafa
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
61d32b3
 
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
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
---
license: creativeml-openrail-m
language:
- ja
tags:
- stable-diffusion
- text-to-image
---

# blue_pencil

<strong>blue_pencil</strong> は、様々なモデルを適当な配合でマージしたモデルです。

有名なモデルをいくつか思い浮かべてください。  
あなたが思い浮かべたモデルは、恐らくこのモデルに含まれています。

このマージモデルの特徴はわかりません。  
いろいろなモデルをマージしてみることが目的なので、質も高くありません。

全てのモデルは [stable-diffusion-webui-model-tookit](https://github.com/arenatemp/stable-diffusion-webui-model-toolkit) を用いて `fp16` にしています。


---

<details open><summary><h2 style="display: inline;"><code>blue_pencil-v2</code> <small>(<code>@20230217</code>)</small></h2></summary>

[AbyssOrangeMix3A1](https://huggingface.co/WarriorMama777/OrangeMixs) をベースに配合しなおしたモデルです。

気持ち程度に自由度が上がっている気がします(同時に破綻率も上がってる気がする)  
基本的には v1 系と同じ雰囲気の絵になる気がします。

以下のモデルが含まれています(順不同)

<details><summary>マージ元モデル一覧</summary>

* [AbyssOrangeMix3A1](https://huggingface.co/WarriorMama777/OrangeMixs)
  * AnythingV3.0
  * ChilloutMix
  * GAPE60
  * Counterfeit2.5
  * Kenshi
* [Evt_M](https://huggingface.co/haor/Evt_M)
  * Evt_V4
    * ACertainty
* [GingerMixR](https://huggingface.co/Hemlok/GingerMix)
  * LimeMixV2
* [Elysium_Anime_V3](https://huggingface.co/hesw23168/SD-Elysium-Model)
* [VaLJMix](https://huggingface.co/Hemlok/VaLMix)
  * pastel-mix
  * ACertainThing
  * basil_mix
  * Counterfeit-V2.5
  * openjourney
* [HD-22](https://www.cognitionai.org/hdhowtogetstarted)
* [7th_anime_v3_testA](https://huggingface.co/syaimu/7th_test)
* [AniReal](https://huggingface.co/Hosioka/AniReal)
* [atwcustom_V4](https://huggingface.co/atsuwo/ATW-custom)
* [Nabylon-v1.2](https://huggingface.co/NegiInNattoMaki/Nabylon-v1.0)
  * AbyssOrangeMix2
  * LonganMix
  * and more
* [TriPhaze_B](https://huggingface.co/Lucetepolis/TriPhaze)
  * ultracolor.v4
  * Counterfeit-V2.5
  * Treebark
* [Balor-V2](https://huggingface.co/ploughB660/Balor-V2)

</details>

### 推奨設定

* VAE
  * [vae-ft-mse-840000-ema-pruned](https://huggingface.co/stabilityai/sd-vae-ft-mse-original)
* Negative Embedding
  * [EasyNegative](https://huggingface.co/datasets/gsdf/EasyNegative)

### 出力例

```
girl, tokyo, scenery
Negative prompt: EasyNegative

Steps: 30, Sampler: DPM++ SDE Karras, CFG scale: 7.5, Seed: 205537258
Size: 768x768, Clip skip: 2
Denoising strength: 0.65, Hires upscale: 2, Hires upscaler: Latent (nearest-exact)
```

![blue_pencil-v2_1](../../resolve/main/images/blue_pencil-v2/1.png)

```
girl, spacesuit, beautiful earth, scenery, on the moon
Negative prompt: EasyNegative

Steps: 50, Sampler: DPM++ SDE Karras, CFG scale: 7.5, Seed: 1069444343
Size: 960x640, Clip skip: 2
Denoising strength: 0.6, Hires upscale: 2, Hires upscaler: Latent (nearest-exact)
```

![blue_pencil-v2_2](../../resolve/main/images/blue_pencil-v2/2.png)

</details>

<details><summary><h2 style="display: inline;"><code>blue_pencil-v1b</code> <small>(<code>@20230212</code>)</small></h2></summary>

`blue_pencil-v1` の [Amalgam_Mix](https://civitai.com/models/4758/amalgammix) の代わりに [Balor-V2](https://huggingface.co/ploughB660/Balor-V2) を階層マージしたモデルです

v1 とはちょっと傾向が違います

### 推奨設定

* VAE
  * [vae-ft-mse-840000-ema-pruned](https://huggingface.co/stabilityai/sd-vae-ft-mse-original)
* Negative Embedding
  * [EasyNegative](https://huggingface.co/datasets/gsdf/EasyNegative)

### 出力例

```
girl, tokyo, scenery
Negative prompt: EasyNegative

Steps: 30, Sampler: DPM++ SDE Karras, CFG scale: 7.5, Seed: 205537258
Size: 768x768, Clip skip: 2
Denoising strength: 0.65, Hires upscale: 2, Hires upscaler: Latent (nearest-exact)
```

![blue_pencil-v1b_1](../../resolve/main/images/blue_pencil-v1b/1.png)

</details>

<details><summary><h2 style="display: inline;"><code>blue_pencil-v1</code> <small>(<code>@20230211</code>)</small></h2></summary>

以下のモデルが含まれています(順不同)

<details><summary>マージ元モデル一覧</summary>

* [Defmix-v1.1](https://huggingface.co/Defpoint/Defmix-v1.0)
  * Counterfeit v1.0
  * Counterfeit v2.0
  * Basil Mix
  * Anything v4.0
* [PastelRainier](https://huggingface.co/Hemlok/RainierMix)
  * ACertainThing
  * Anything-V4.5
  * Counterfeit-V2.0
  * Evt_V4-preview
  * basil_mix
  * pastel-mix
* [GingerMixR](https://huggingface.co/Hemlok/GingerMix)
  * LimeMixV2
* [Elysium_Anime_V3](https://huggingface.co/hesw23168/SD-Elysium-Model)
* [SukiyakiMix-v1.0](https://huggingface.co/Vsukiyaki/SukiyakiMix-v1.0)
  * pastel-mix
  * AbyssOrangeMix2
* [HD-20](https://www.cognitionai.org/hdhowtogetstarted)
* [7th_anime_v3_testA](https://huggingface.co/syaimu/7th_test)
* [AniReal](https://huggingface.co/Hosioka/AniReal)
* [TriPhaze_B](https://huggingface.co/Lucetepolis/TriPhaze)
  * ultracolor.v4
  * Counterfeit-V2.5
  * Treebark
* [Nabylon-v1.2](https://huggingface.co/NegiInNattoMaki/Nabylon-v1.0)
  * AbyssOrangeMix2
  * LonganMix
  * and more
* [atwcustom_V4](https://huggingface.co/atsuwo/ATW-custom)
* [Amalgam_Mix](https://civitai.com/models/4758/amalgammix)

</details>

### 推奨設定

* VAE
  * [vae-ft-mse-840000-ema-pruned](https://huggingface.co/stabilityai/sd-vae-ft-mse-original)
* Negative Embedding
  * [EasyNegative](https://huggingface.co/datasets/gsdf/EasyNegative)

### 出力例

#### 1

```
girl, tokyo, scenery
Negative prompt: EasyNegative

Steps: 30, Sampler: DPM++ SDE Karras, CFG scale: 7.5, Seed: 2526423076
Size: 768x768, Clip skip: 2
```

![blue_pencil-v1_1-1](../../resolve/main/images/blue_pencil-v1/1-1.png)

##### Hires. fix

```
Denoising strength: 0.6, Hires upscale: 2, Hires upscaler: Latent (nearest-exact)
```

![blue_pencil-v1_1-2](../../resolve/main/images/blue_pencil-v1/1-2.png)

#### 2

```
girl, early teen, kimono, sakura, particles
Negative prompt: EasyNegative

Steps: 20, Sampler: DPM++ SDE Karras, CFG scale: 7.5, Seed: 4036639388,
Size: 512x768, Clip skip: 2
```

![blue_pencil-v1_2-1](../../resolve/main/images/blue_pencil-v1/2-1.png)

##### Hires. fix

```
Denoising strength: 0.62, Hires upscale: 2, Hires upscaler: Latent (nearest-exact)
```

![blue_pencil-v1_2-2](../../resolve/main/images/blue_pencil-v1/2-2.png)

#### 3

```
girl, early teen, t-shirt, pants, from behind, landscape, scenery, apocalyptic
Negative prompt: EasyNegative

Steps: 40, Sampler: DPM++ SDE Karras, CFG scale: 7.5, Seed: 748447692,
Size: 768x512, Clip skip: 2
```

![blue_pencil-v1_3](../../resolve/main/images/blue_pencil-v1/3.png)

</details>