Skip to content

Commit 67727d9

Browse files
mintlify[bot]mintlify[bot]
authored
Add NewBie-image Exp0.1 tutorials (#666)
* Update tutorials/image/newbie-image/newbie-image-exp-0-1.mdx Co-Authored-By: mintlify[bot] <109931778+mintlify[bot]@users.noreply.github.com> * Update zh-CN/tutorials/image/newbie-image/newbie-image-exp-0-1.mdx Co-Authored-By: mintlify[bot] <109931778+mintlify[bot]@users.noreply.github.com> * Update docs.json Co-Authored-By: mintlify[bot] <109931778+mintlify[bot]@users.noreply.github.com> * Update docs.json Co-Authored-By: mintlify[bot] <109931778+mintlify[bot]@users.noreply.github.com> * Update tutorials/image/newbie-image/newbie-image-exp-0-1.mdx Co-Authored-By: mintlify[bot] <109931778+mintlify[bot]@users.noreply.github.com> * Update zh-CN/tutorials/image/newbie-image/newbie-image-exp-0-1.mdx Co-Authored-By: mintlify[bot] <109931778+mintlify[bot]@users.noreply.github.com> --------- Co-authored-by: mintlify[bot] <109931778+mintlify[bot]@users.noreply.github.com>
1 parent 033c220 commit 67727d9

File tree

3 files changed

+268
-0
lines changed

3 files changed

+268
-0
lines changed

‎docs.json‎

Lines changed: 12 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -164,6 +164,12 @@
164164
"tutorials/image/ovis/ovis-image"
165165
]
166166
},
167+
{
168+
"group": "NewBie-image",
169+
"pages": [
170+
"tutorials/image/newbie-image/newbie-image-exp-0-1"
171+
]
172+
},
167173
"tutorials/image/cosmos/cosmos-predict2-t2i",
168174
"tutorials/image/omnigen/omnigen2"
169175
]
@@ -809,6 +815,12 @@
809815
"zh-CN/tutorials/image/ovis/ovis-image"
810816
]
811817
},
818+
{
819+
"group": "NewBie-image",
820+
"pages": [
821+
"zh-CN/tutorials/image/newbie-image/newbie-image-exp-0-1"
822+
]
823+
},
812824
"zh-CN/tutorials/image/cosmos/cosmos-predict2-t2i",
813825
"zh-CN/tutorials/image/omnigen/omnigen2"
814826
]

‎tutorials/image/newbie-image/newbie-image-exp-0-1.mdx‎

Lines changed: 128 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,128 @@
1+
---
2+
title: "ComfyUI NewBie-image-Exp0.1 Workflow Example"
3+
description: "NewBie-image-Exp0.1 is a 3.5B parameter anime-style text-to-image generation model based on Next-DiT architecture, optimized for high-quality anime image generation with XML structured prompts."
4+
sidebarTitle: "NewBie-image-Exp0.1"
5+
---
6+
7+
import UpdateReminder from '/snippets/tutorials/update-reminder.mdx'
8+
9+
**NewBie-image-Exp0.1** is a 3.5B parameter DiT model developed by NewBieAI Lab for anime-style text-to-image generation. Built on the Next-DiT architecture, it delivers remarkably detailed and visually striking anime images.
10+
11+
**Key Features**:
12+
- **3.5B Parameter Model**: Efficient yet powerful model size for high-quality anime generation
13+
- **Next-DiT Architecture**: Based on research from the Lumina architecture with a newly designed NewBie architecture
14+
- **Dual Text Encoders**: Uses Gemma3-4B-it as primary encoder with Jina CLIP v2 for improved prompt understanding
15+
- **FLUX VAE**: Utilizes FLUX.1-dev 16-channel VAE for richer colors and finer texture details
16+
- **XML Structured Prompts**: Supports XML format for better attention binding and attribute disentanglement
17+
18+
**Related Links**:
19+
- [GitHub](https://github.com/NewBieAI-Lab/NewBie-image-Exp0.1)
20+
- [Hugging Face](https://huggingface.co/NewBie-AI/NewBie-image-Exp0.1)
21+
- [Getting Started Guide](https://ai.feishu.cn/wiki/NZl9wm7V1iuNzmkRKCUcb1USnsh)
22+
23+
## NewBie-image text-to-image workflow
24+
25+
<a className="prose" target='_blank' href="https://raw.githubusercontent.com/Comfy-Org/workflow_templates/refs/heads/main/templates/image_newbieimage_exp0_1-t2i.json" style={{ display: 'inline-block', backgroundColor: '#0078D6', color: '#ffffff', padding: '10px 20px', borderRadius: '8px', borderColor: "transparent", textDecoration: 'none', fontWeight: 'bold', marginRight: '10px'}}>
26+
<p className="prose" style={{ margin: 0, fontSize: "0.8rem" }}>Download JSON Workflow File</p>
27+
</a>
28+
29+
<a className="prose" target='_blank' href="https://cloud.comfy.org/?template=image_newbieimage_exp0_1-t2i&utm_source=docs" style={{ display: 'inline-block', backgroundColor: '#28a745', color: '#ffffff', padding: '10px 20px', borderRadius: '8px', borderColor: "transparent", textDecoration: 'none', fontWeight: 'bold'}}>
30+
<p className="prose" style={{ margin: 0, fontSize: "0.8rem" }}>Run on ComfyUI Cloud</p>
31+
</a>
32+
33+
<UpdateReminder />
34+
35+
## Model links
36+
37+
**text_encoders**
38+
39+
- [gemma_3_4b_it_bf16.safetensors](https://huggingface.co/Comfy-Org/NewBie-image-Exp0.1_repackaged/resolve/main/split_files/text_encoders/gemma_3_4b_it_bf16.safetensors)
40+
- [jina_clip_v2_bf16.safetensors](https://huggingface.co/Comfy-Org/NewBie-image-Exp0.1_repackaged/resolve/main/split_files/text_encoders/jina_clip_v2_bf16.safetensors)
41+
42+
**diffusion_models**
43+
44+
- [NewBie-Image-Exp0.1-bf16.safetensors](https://huggingface.co/Comfy-Org/NewBie-image-Exp0.1_repackaged/resolve/main/split_files/diffusion_models/NewBie-Image-Exp0.1-bf16.safetensors)
45+
46+
**vae**
47+
48+
- [ae.safetensors](https://huggingface.co/Comfy-Org/z_image_turbo/resolve/main/split_files/vae/ae.safetensors)
49+
50+
**Model Storage Location**
51+
52+
```
53+
ComfyUI/
54+
├── models/
55+
│ ├── text_encoders/
56+
│ │ ├── gemma_3_4b_it_bf16.safetensors
57+
│ │ └── jina_clip_v2_bf16.safetensors
58+
│ ├── diffusion_models/
59+
│ │ └── NewBie-Image-Exp0.1-bf16.safetensors
60+
│ └── vae/
61+
│ └── ae.safetensors
62+
```
63+
64+
## Prompt format
65+
66+
NewBie-image is an anime image generation model optimized for character generation. It uses XML structured prompts for training, where each `<>` tag defines a category (like `<appearance>`, `<clothing>`) and `</>` closes it. The tags inside are standard Danbooru tags. This structure enables precise control over multi-character scenes with better attribute binding.
67+
68+
For the complete prompt writing guide, see the [official documentation](https://ai.feishu.cn/wiki/NZl9wm7V1iuNzmkRKCUcb1USnsh).
69+
70+
NewBie-image-Exp0.1 supports three prompt formats:
71+
- **Natural language**: Standard text descriptions
72+
- **Tags**: Danbooru-style tags
73+
- **XML structured format**: Recommended for multi-character scenes
74+
75+
### XML structured prompt
76+
77+
For multi-character scenes, using XML structured prompts typically leads to more accurate image generation results with better attention binding and attribute disentanglement.
78+
79+
```xml
80+
<character_1>
81+
<n>$character_1$</n>
82+
<gender>1girl</gender>
83+
<appearance>chibi, red_eyes, blue_hair, long_hair, hair_between_eyes, head_tilt, tareme, closed_mouth</appearance>
84+
<clothing>school_uniform, serafuku, white_sailor_collar, white_shirt, short_sleeves, red_neckerchief, bow, blue_skirt, miniskirt, pleated_skirt, blue_hat, mini_hat, thighhighs, grey_thighhighs, black_shoes, mary_janes</clothing>
85+
<expression>happy, smile</expression>
86+
<action>standing, holding, holding_briefcase</action>
87+
<position>center_left</position>
88+
</character_1>
89+
90+
<character_2>
91+
<n>$character_2$</n>
92+
<gender>1girl</gender>
93+
<appearance>chibi, red_eyes, pink_hair, long_hair, very_long_hair, multi-tied_hair, open_mouth</appearance>
94+
<clothing>school_uniform, serafuku, white_sailor_collar, white_shirt, short_sleeves, red_neckerchief, bow, red_skirt, miniskirt, pleated_skirt, hair_bow, multiple_hair_bows, white_bow, ribbon_trim, ribbon-trimmed_bow, white_thighhighs, black_shoes, mary_janes, bow_legwear, bare_arms</clothing>
95+
<expression>happy, smile</expression>
96+
<action>standing, holding, holding_briefcase, waving</action>
97+
<position>center_right</position>
98+
</character_2>
99+
100+
<general_tags>
101+
<count>2girls, multiple_girls</count>
102+
<style>anime_style, digital_art</style>
103+
<background>white_background, simple_background</background>
104+
<atmosphere>cheerful</atmosphere>
105+
<quality>high_resolution, detailed</quality>
106+
<objects>briefcase</objects>
107+
<other>alternate_costume</other>
108+
</general_tags>
109+
```
110+
111+
### XML tag reference
112+
113+
| Tag | Description |
114+
|-----|-------------|
115+
| `<n>` | Character name/identifier |
116+
| `<gender>` | Character gender (1girl, 1boy, etc.) |
117+
| `<appearance>` | Physical features (hair, eyes, body type) |
118+
| `<clothing>` | Outfit and accessories |
119+
| `<expression>` | Facial expression |
120+
| `<action>` | Pose and actions |
121+
| `<position>` | Position in the image |
122+
| `<count>` | Number of characters |
123+
| `<style>` | Art style |
124+
| `<background>` | Background description |
125+
| `<atmosphere>` | Overall mood |
126+
| `<quality>` | Quality tags |
127+
| `<objects>` | Objects in the scene |
128+
| `<other>` | Additional tags |

‎zh-CN/tutorials/image/newbie-image/newbie-image-exp-0-1.mdx‎

Lines changed: 128 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,128 @@
1+
---
2+
title: "ComfyUI NewBie-image-Exp0.1 工作流示例"
3+
description: "NewBie-image-Exp0.1 是一个基于 Next-DiT 架构的 35 亿参数动漫风格文生图模型，针对高质量动漫图像生成进行了优化，支持 XML 结构化提示词。"
4+
sidebarTitle: "NewBie-image-Exp0.1"
5+
---
6+
7+
import UpdateReminder from '/snippets/tutorials/update-reminder.mdx'
8+
9+
**NewBie-image-Exp0.1** 是由 NewBieAI Lab 开发的 35 亿参数 DiT 模型，专为动漫风格文生图设计。基于 Next-DiT 架构构建，能够生成细节丰富、视觉效果出色的动漫图像。
10+
11+
**核心特性**：
12+
- **35 亿参数模型**：高效且强大的模型规模，适合高质量动漫生成
13+
- **Next-DiT 架构**：基于 Lumina 架构研究，采用全新设计的 NewBie 架构
14+
- **双文本编码器**：使用 Gemma3-4B-it 作为主编码器，配合 Jina CLIP v2 提升提示词理解能力
15+
- **FLUX VAE**：采用 FLUX.1-dev 16 通道 VAE，呈现更丰富的色彩和更精细的纹理细节
16+
- **XML 结构化提示词**：支持 XML 格式，实现更好的注意力绑定和属性解耦
17+
18+
**相关链接**：
19+
- [GitHub](https://github.com/NewBieAI-Lab/NewBie-image-Exp0.1)
20+
- [Hugging Face](https://huggingface.co/NewBie-AI/NewBie-image-Exp0.1)
21+
- [入门指南](https://ai.feishu.cn/wiki/P3sgwUUjWih8ZWkpr0WcwXSMnTb)
22+
23+
## NewBie-image 文生图工作流
24+
25+
<a className="prose" target='_blank' href="https://raw.githubusercontent.com/Comfy-Org/workflow_templates/refs/heads/main/templates/image_newbieimage_exp0_1-t2i.json" style={{ display: 'inline-block', backgroundColor: '#0078D6', color: '#ffffff', padding: '10px 20px', borderRadius: '8px', borderColor: "transparent", textDecoration: 'none', fontWeight: 'bold', marginRight: '10px'}}>
26+
<p className="prose" style={{ margin: 0, fontSize: "0.8rem" }}>下载 JSON 工作流文件</p>
27+
</a>
28+
29+
<a className="prose" target='_blank' href="https://cloud.comfy.org/?template=image_newbieimage_exp0_1-t2i&utm_source=docs" style={{ display: 'inline-block', backgroundColor: '#28a745', color: '#ffffff', padding: '10px 20px', borderRadius: '8px', borderColor: "transparent", textDecoration: 'none', fontWeight: 'bold'}}>
30+
<p className="prose" style={{ margin: 0, fontSize: "0.8rem" }}>在 ComfyUI Cloud 上运行</p>
31+
</a>
32+
33+
<UpdateReminder />
34+
35+
## 模型下载链接
36+
37+
**text_encoders**
38+
39+
- [gemma_3_4b_it_bf16.safetensors](https://huggingface.co/Comfy-Org/NewBie-image-Exp0.1_repackaged/resolve/main/split_files/text_encoders/gemma_3_4b_it_bf16.safetensors)
40+
- [jina_clip_v2_bf16.safetensors](https://huggingface.co/Comfy-Org/NewBie-image-Exp0.1_repackaged/resolve/main/split_files/text_encoders/jina_clip_v2_bf16.safetensors)
41+
42+
**diffusion_models**
43+
44+
- [NewBie-Image-Exp0.1-bf16.safetensors](https://huggingface.co/Comfy-Org/NewBie-image-Exp0.1_repackaged/resolve/main/split_files/diffusion_models/NewBie-Image-Exp0.1-bf16.safetensors)
45+
46+
**vae**
47+
48+
- [ae.safetensors](https://huggingface.co/Comfy-Org/z_image_turbo/resolve/main/split_files/vae/ae.safetensors)
49+
50+
**模型存放位置**
51+
52+
```
53+
ComfyUI/
54+
├── models/
55+
│ ├── text_encoders/
56+
│ │ ├── gemma_3_4b_it_bf16.safetensors
57+
│ │ └── jina_clip_v2_bf16.safetensors
58+
│ ├── diffusion_models/
59+
│ │ └── NewBie-Image-Exp0.1-bf16.safetensors
60+
│ └── vae/
61+
│ └── ae.safetensors
62+
```
63+
64+
## 提示词格式
65+
66+
NewBie-image 是一个针对角色生成优化的动漫图像生成模型。它使用 XML 结构化提示词进行训练，每个 `<>` 标签定义一个类别（如 `<appearance>`、`<clothing>`），`</>` 作为结束标记。标签内部使用标准的 Danbooru 标签。这种结构能够精确控制多角色场景，实现更好的属性绑定。
67+
68+
完整的提示词编写指南请参阅[官方文档](https://ai.feishu.cn/wiki/NZl9wm7V1iuNzmkRKCUcb1USnsh)。
69+
70+
NewBie-image-Exp0.1 支持三种提示词格式：
71+
- **自然语言**：标准文本描述
72+
- **标签**：Danbooru 风格标签
73+
- **XML 结构化格式**：推荐用于多角色场景
74+
75+
### XML 结构化提示词
76+
77+
对于多角色场景，使用 XML 结构化提示词通常能获得更准确的图像生成结果，具有更好的注意力绑定和属性解耦效果。
78+
79+
```xml
80+
<character_1>
81+
<n>$character_1$</n>
82+
<gender>1girl</gender>
83+
<appearance>chibi, red_eyes, blue_hair, long_hair, hair_between_eyes, head_tilt, tareme, closed_mouth</appearance>
84+
<clothing>school_uniform, serafuku, white_sailor_collar, white_shirt, short_sleeves, red_neckerchief, bow, blue_skirt, miniskirt, pleated_skirt, blue_hat, mini_hat, thighhighs, grey_thighhighs, black_shoes, mary_janes</clothing>
85+
<expression>happy, smile</expression>
86+
<action>standing, holding, holding_briefcase</action>
87+
<position>center_left</position>
88+
</character_1>
89+
90+
<character_2>
91+
<n>$character_2$</n>
92+
<gender>1girl</gender>
93+
<appearance>chibi, red_eyes, pink_hair, long_hair, very_long_hair, multi-tied_hair, open_mouth</appearance>
94+
<clothing>school_uniform, serafuku, white_sailor_collar, white_shirt, short_sleeves, red_neckerchief, bow, red_skirt, miniskirt, pleated_skirt, hair_bow, multiple_hair_bows, white_bow, ribbon_trim, ribbon-trimmed_bow, white_thighhighs, black_shoes, mary_janes, bow_legwear, bare_arms</clothing>
95+
<expression>happy, smile</expression>
96+
<action>standing, holding, holding_briefcase, waving</action>
97+
<position>center_right</position>
98+
</character_2>
99+
100+
<general_tags>
101+
<count>2girls, multiple_girls</count>
102+
<style>anime_style, digital_art</style>
103+
<background>white_background, simple_background</background>
104+
<atmosphere>cheerful</atmosphere>
105+
<quality>high_resolution, detailed</quality>
106+
<objects>briefcase</objects>
107+
<other>alternate_costume</other>
108+
</general_tags>
109+
```
110+
111+
### XML 标签参考
112+
113+
| 标签 | 描述 |
114+
|-----|-------------|
115+
| `<n>` | 角色名称/标识符 |
116+
| `<gender>` | 角色性别（1girl、1boy 等） |
117+
| `<appearance>` | 外貌特征（头发、眼睛、体型） |
118+
| `<clothing>` | 服装和配饰 |
119+
| `<expression>` | 面部表情 |
120+
| `<action>` | 姿势和动作 |
121+
| `<position>` | 图像中的位置 |
122+
| `<count>` | 角色数量 |
123+
| `<style>` | 艺术风格 |
124+
| `<background>` | 背景描述 |
125+
| `<atmosphere>` | 整体氛围 |
126+
| `<quality>` | 质量标签 |
127+
| `<objects>` | 场景中的物品 |
128+
| `<other>` | 其他标签 |

0 commit comments

Comments
 (0)