Update README.md
Browse files
README.md
CHANGED
@@ -4,7 +4,7 @@ emoji: 🎬
|
|
4 |
colorFrom: indigo
|
5 |
colorTo: purple
|
6 |
sdk: gradio
|
7 |
-
sdk_version: 5.
|
8 |
app_file: app.py
|
9 |
pinned: false
|
10 |
short_description: NSFW Uncensored text/Image to image for AI Limits
|
@@ -14,5 +14,105 @@ models:
|
|
14 |
- VIDraft/Gemma-3-R1984-12B
|
15 |
- VIDraft/Gemma-3-R1984-4B
|
16 |
---
|
|
|
17 |
|
18 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
4 |
colorFrom: indigo
|
5 |
colorTo: purple
|
6 |
sdk: gradio
|
7 |
+
sdk_version: 5.35.0
|
8 |
app_file: app.py
|
9 |
pinned: false
|
10 |
short_description: NSFW Uncensored text/Image to image for AI Limits
|
|
|
14 |
- VIDraft/Gemma-3-R1984-12B
|
15 |
- VIDraft/Gemma-3-R1984-4B
|
16 |
---
|
17 |
+
Looking at your code, I'll create an English description for your NSFW image generation application:
|
18 |
|
19 |
+
---
|
20 |
+
|
21 |
+
# NSFW Uncensored AI Image Generator
|
22 |
+
|
23 |
+
## Overview
|
24 |
+
This is an advanced AI-powered image generation platform that specializes in creating uncensored adult content using Stable Diffusion XL technology. The application features both text-to-image and image-to-image generation capabilities with multilingual support.
|
25 |
+
|
26 |
+
## Key Features
|
27 |
+
|
28 |
+
### 🌐 Multilingual Support
|
29 |
+
- **Universal Language Input**: Enter prompts in any language (Korean, English, Japanese, Chinese, etc.)
|
30 |
+
- **Automatic Translation**: Non-English prompts are automatically translated to English using Cohere AI
|
31 |
+
- **Smart Detection**: Automatically detects Korean (한글), Japanese (日本語), Chinese (中文), and other non-English text
|
32 |
+
|
33 |
+
### 🎨 Generation Modes
|
34 |
+
|
35 |
+
#### Text-to-Image
|
36 |
+
- Create images from text descriptions
|
37 |
+
- Keyword boost feature for optimal prompt generation
|
38 |
+
- Random prompt suggestions for inspiration
|
39 |
+
- Advanced controls for fine-tuning results
|
40 |
+
|
41 |
+
#### Image-to-Image
|
42 |
+
- Transform existing images with text prompts
|
43 |
+
- Adjustable transformation strength
|
44 |
+
- Preserve original image characteristics while applying new styles
|
45 |
+
|
46 |
+
### 🛠️ Technical Specifications
|
47 |
+
|
48 |
+
**Model**: Heartsync/NSFW-Uncensored (Stable Diffusion XL)
|
49 |
+
- Optimized for adult content generation
|
50 |
+
- High-quality output up to 1216x1216 pixels
|
51 |
+
- FP16 precision for efficient VRAM usage
|
52 |
+
|
53 |
+
**Advanced Settings**:
|
54 |
+
- **Guidance Scale**: 0-20 (default: 7)
|
55 |
+
- **Inference Steps**: 1-50 steps
|
56 |
+
- **Seed Control**: Manual or randomized
|
57 |
+
- **Negative Prompts**: Specify what to avoid
|
58 |
+
- **Transformation Strength**: 0-1 for image-to-image
|
59 |
+
|
60 |
+
### 🚀 Special Features
|
61 |
+
|
62 |
+
**BOOST Button**:
|
63 |
+
- Generates optimized NSFW prompts from simple keywords
|
64 |
+
- Uses Cohere AI to create detailed, explicit descriptions
|
65 |
+
- Ensures high-quality, uncensored content generation
|
66 |
+
|
67 |
+
**RANDOM Button**:
|
68 |
+
- Selects from curated examples of effective prompts
|
69 |
+
- Great for inspiration and understanding prompt structure
|
70 |
+
|
71 |
+
### 💡 Usage Tips
|
72 |
+
|
73 |
+
1. **For Best Results**:
|
74 |
+
- Use descriptive, detailed prompts
|
75 |
+
- Combine physical descriptions with actions and atmosphere
|
76 |
+
- Leverage the BOOST feature for keyword expansion
|
77 |
+
|
78 |
+
2. **Language Support**:
|
79 |
+
- Write prompts in your native language
|
80 |
+
- Automatic translation ensures accurate generation
|
81 |
+
- Both positive and negative prompts support all languages
|
82 |
+
|
83 |
+
3. **Image-to-Image Tips**:
|
84 |
+
- Start with lower strength (0.5-0.7) to preserve original features
|
85 |
+
- Use higher strength (0.8-1.0) for dramatic transformations
|
86 |
+
- Combine with descriptive prompts for best results
|
87 |
+
|
88 |
+
### ⚡ Performance
|
89 |
+
|
90 |
+
- GPU-accelerated generation using CUDA
|
91 |
+
- Optimized for Hugging Face Spaces infrastructure
|
92 |
+
- Typical generation time: 10-30 seconds per image
|
93 |
+
- Queue system for handling multiple requests
|
94 |
+
|
95 |
+
### 🔧 API Requirements
|
96 |
+
|
97 |
+
- **Cohere API**: Required for translation and prompt generation
|
98 |
+
- Set `COH_API` environment variable with your Cohere API key
|
99 |
+
- Uses Command-R Plus model for language processing
|
100 |
+
|
101 |
+
### 📝 Example Prompts
|
102 |
+
|
103 |
+
**Simple**: "A romantic couple in an intimate moment"
|
104 |
+
|
105 |
+
**Detailed**: "A shy college girl with glasses and tight plaid skirt nervously approaches her professor in a dimly lit classroom"
|
106 |
+
|
107 |
+
**Atmospheric**: "Moody mature anime scene of two lovers under neon rain, sensual atmosphere"
|
108 |
+
|
109 |
+
### ⚠️ Important Notes
|
110 |
+
|
111 |
+
- This tool is designed for adult content creation
|
112 |
+
- All generated content is uncensored
|
113 |
+
- Users must be 18+ to use this application
|
114 |
+
- Please use responsibly and respect content policies
|
115 |
+
|
116 |
+
---
|
117 |
+
|
118 |
+
This platform pushes the boundaries of AI image generation, providing creators with powerful tools for generating unrestricted adult content across multiple languages and styles.
|