Update README.md
Browse files
README.md
CHANGED
@@ -17,28 +17,34 @@ models:
|
|
17 |
- VIDraft/Gemma-3-R1984-4B
|
18 |
---
|
19 |
|
20 |
-
ChartGPT : AI-Powered Multi Diagram Generator
|
21 |
VIDEO Link: https://www.youtube.com/watch?v=erof86WoNMY
|
22 |
|
23 |
<div align="center">
|
24 |
-
<h3>๐จ AI Service that Generates Professional Diagrams from Natural Language</h3>
|
25 |
-
<p>Based LLM 'Gemma-3-R1984-27B' Model. Powered by VIDraftโก</p>
|
26 |
-
<p><em>Ranked 2nd on Google DeepMind
|
27 |
</div>
|
28 |
|
29 |
-
Service Link:
|
30 |
|
31 |
๐ Overview
|
32 |
-
ChartGPT is an AI-based service that automatically generates professional diagrams from Korean and English prompts. Without needing to learn complex diagramming tools, simply describe what you want and AI will automatically generate structured diagrams.
|
33 |
|
34 |
๐ Core Technology
|
35 |
-
|
36 |
|
37 |
-
Achieved 2nd place on Google DeepMind FACTS Grounding Leaderboard
|
38 |
State-of-the-art language model with exceptional factuality and accuracy
|
39 |
Outstanding performance in both Korean and English
|
40 |
Optimized for complex structuring tasks
|
41 |
|
|
|
|
|
|
|
|
|
|
|
|
|
42 |
MCP (Model Context Protocol) Support
|
43 |
|
44 |
Standardized AI Integration: Full support for Anthropic's MCP protocol
|
@@ -54,19 +60,29 @@ Extensible Architecture: Operates as an MCP server for easy connection with othe
|
|
54 |
๐ Process Flow: Represent workflows or decision-making processes sequentially
|
55 |
๐ WBS Diagram: Display projects broken down hierarchically
|
56 |
|
57 |
-
2.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
58 |
|
59 |
Generate diagrams from natural language prompts
|
60 |
-
|
61 |
Automatic structured JSON generation
|
|
|
62 |
|
63 |
-
|
64 |
|
65 |
Reflect latest information through real-time web search
|
66 |
Generate more accurate and richer content
|
67 |
Optional activation
|
68 |
|
69 |
-
|
70 |
|
71 |
PNG format (high-quality images)
|
72 |
SVG format (vector graphics, infinitely scalable)
|
@@ -81,57 +97,59 @@ Multimodal AI Pipeline: Automate text analysis โ diagram generation
|
|
81 |
|
82 |
๐ How to Use
|
83 |
Using the AI Assistant Tab
|
|
|
84 |
|
85 |
-
|
86 |
-
|
87 |
-
|
88 |
-
Example: "Create a flowchart showing the machine learning process"
|
89 |
-
|
90 |
|
91 |
-
Select
|
92 |
|
93 |
-
|
|
|
94 |
|
95 |
-
|
96 |
-
Use Brave Search (Optional)
|
97 |
|
98 |
Activate checkbox when latest information is needed
|
99 |
-
Web search results will be reflected in
|
100 |
-
|
101 |
-
|
102 |
-
Click Generate Button
|
103 |
-
|
104 |
-
AI automatically generates JSON structure
|
105 |
-
Diagram image automatically rendered
|
106 |
|
|
|
107 |
|
|
|
|
|
108 |
|
109 |
Direct JSON Input
|
|
|
110 |
Fine-tune by directly editing JSON in each diagram type tab
|
|
|
111 |
๐ฏ Use Cases
|
112 |
Business
|
113 |
|
114 |
Project planning (WBS)
|
115 |
Business process documentation (Process Flow)
|
116 |
Organizational chart creation (Radial Diagram)
|
|
|
|
|
117 |
|
118 |
Education
|
119 |
|
120 |
-
Learning content organization (Concept Map)
|
121 |
Course explanatory materials (Synoptic Chart)
|
122 |
Research topic structuring (Concept Map)
|
|
|
123 |
|
124 |
Technical Documentation
|
125 |
|
126 |
System architecture diagramming
|
127 |
API flow explanation
|
128 |
Data processing pipeline visualization
|
|
|
|
|
129 |
|
130 |
๐ง Advanced Configuration
|
131 |
Customizing JSON Structure
|
132 |
Each diagram type has its unique JSON structure:
|
133 |
Concept Map Example:
|
134 |
-
|
135 |
json{
|
136 |
"central_node": "Artificial Intelligence",
|
137 |
"nodes": [
|
@@ -143,34 +161,43 @@ json{
|
|
143 |
}
|
144 |
]
|
145 |
}
|
146 |
-
|
147 |
Prompt Writing Tips
|
148 |
|
149 |
Be Specific: "AI classification system" โ "AI classification system including deep learning, machine learning, and natural language processing"
|
150 |
Specify Relationships: "A is part of B", "X leads to Y", etc.
|
151 |
Describe Hierarchical Structure: "A at the top level, with B and C below", etc.
|
|
|
152 |
|
153 |
๐ก Model Performance
|
154 |
Gemma-3-R1984-27B Features
|
155 |
|
156 |
-
High Factuality
|
157 |
Structuring Ability: Accurately understands complex relationships and hierarchical structures
|
158 |
Multilingual Support: Ensures equal quality in both Korean and English
|
159 |
Context Understanding: Accurately grasps prompt intent
|
160 |
|
|
|
|
|
|
|
|
|
|
|
|
|
161 |
๐ ๏ธ Tech Stack
|
162 |
|
163 |
Frontend: Gradio (with built-in MCP server)
|
164 |
-
AI
|
|
|
|
|
|
|
|
|
|
|
165 |
Diagram Engine: Graphviz
|
166 |
Search Engine: Brave Search API
|
167 |
Protocol: MCP (Model Context Protocol)
|
168 |
Language: Python
|
169 |
|
170 |
-
|
171 |
<div align="center">
|
172 |
<p>Made with โค๏ธ by VIDraft Team</p>
|
173 |
<p><strong>Powered by VIDraftโก</strong></p>
|
174 |
<p><em>MCP-enabled for seamless AI integration</em></p>
|
175 |
-
</div>
|
176 |
-
|
|
|
17 |
- VIDraft/Gemma-3-R1984-4B
|
18 |
---
|
19 |
|
20 |
+
ChartGPT : AI-Powered Multi Diagram & Design Generator
|
21 |
VIDEO Link: https://www.youtube.com/watch?v=erof86WoNMY
|
22 |
|
23 |
<div align="center">
|
24 |
+
<h3>๐จ AI Service that Generates Professional Diagrams and Visual Designs from Natural Language</h3>
|
25 |
+
<p>Based on LLM 'Gemma-3-R1984-27B' Model & FLUX Image Generation Model. Powered by VIDraftโก</p>
|
26 |
+
<p><em>Ranked 2nd on Google DeepMind FACTS Grounding Leaderboard</em></p>
|
27 |
</div>
|
28 |
|
29 |
+
Service Link: https://huggingface.co/spaces/openfree/Chart-GPT
|
30 |
|
31 |
๐ Overview
|
32 |
+
ChartGPT is an AI-based service that automatically generates professional diagrams and visual designs from Korean and English prompts. Without needing to learn complex diagramming tools, simply describe what you want and AI will automatically generate structured diagrams or beautiful visual designs.
|
33 |
|
34 |
๐ Core Technology
|
35 |
+
Base LLM: Gemma-3-R1984-27B
|
36 |
|
37 |
+
Achieved 2nd place on Google DeepMind FACTS Grounding Leaderboard(https://huggingface.co/spaces/MaziyarPanahi/FACTS-Leaderboard)
|
38 |
State-of-the-art language model with exceptional factuality and accuracy
|
39 |
Outstanding performance in both Korean and English
|
40 |
Optimized for complex structuring tasks
|
41 |
|
42 |
+
Image Generation: FLUX.1-schnell
|
43 |
+
|
44 |
+
High-quality image generation based on latest diffusion models
|
45 |
+
Fast generation speed with high quality
|
46 |
+
Support for various styles and concepts
|
47 |
+
|
48 |
MCP (Model Context Protocol) Support
|
49 |
|
50 |
Standardized AI Integration: Full support for Anthropic's MCP protocol
|
|
|
60 |
๐ Process Flow: Represent workflows or decision-making processes sequentially
|
61 |
๐ WBS Diagram: Display projects broken down hierarchically
|
62 |
|
63 |
+
2. Support for 6 Visual Design Types (NEW)
|
64 |
+
|
65 |
+
๐จ Product Design: Industrial design concepts and product sketches
|
66 |
+
๐ง Mindmap: Visually rich thought maps
|
67 |
+
๐ฑ Mockup: UI/UX wireframes and app designs
|
68 |
+
๐ Infographic: Data visualization and information design
|
69 |
+
๐ Diagram: Business workflows and process visualization
|
70 |
+
๐ Flowchart: Decision flows and process charts
|
71 |
+
|
72 |
+
3. AI-Based Automatic Generation
|
73 |
|
74 |
Generate diagrams from natural language prompts
|
75 |
+
Automatic translation from Korean to English (for Design Types)
|
76 |
Automatic structured JSON generation
|
77 |
+
Optimized prompt templates for each design type
|
78 |
|
79 |
+
4. Brave Search Integration
|
80 |
|
81 |
Reflect latest information through real-time web search
|
82 |
Generate more accurate and richer content
|
83 |
Optional activation
|
84 |
|
85 |
+
5. Flexible Output Formats
|
86 |
|
87 |
PNG format (high-quality images)
|
88 |
SVG format (vector graphics, infinitely scalable)
|
|
|
97 |
|
98 |
๐ How to Use
|
99 |
Using the AI Assistant Tab
|
100 |
+
1. Enter Prompt
|
101 |
|
102 |
+
Describe the diagram or design you want in natural language
|
103 |
+
Diagram example: "Create a flowchart showing the machine learning process"
|
104 |
+
Design example: "smartphone banking app design"
|
|
|
|
|
105 |
|
106 |
+
2. Select Type
|
107 |
|
108 |
+
Diagram Type: Generate structured diagrams (JSON-based)
|
109 |
+
Design Type: Generate visual designs (FLUX-based, English only)
|
110 |
|
111 |
+
3. Use Brave Search (Optional)
|
|
|
112 |
|
113 |
Activate checkbox when latest information is needed
|
114 |
+
Web search results will be reflected in generation
|
|
|
|
|
|
|
|
|
|
|
|
|
115 |
|
116 |
+
4. Click Generate Button
|
117 |
|
118 |
+
AI automatically generates structure
|
119 |
+
Diagram or design image automatically rendered
|
120 |
|
121 |
Direct JSON Input
|
122 |
+
|
123 |
Fine-tune by directly editing JSON in each diagram type tab
|
124 |
+
|
125 |
๐ฏ Use Cases
|
126 |
Business
|
127 |
|
128 |
Project planning (WBS)
|
129 |
Business process documentation (Process Flow)
|
130 |
Organizational chart creation (Radial Diagram)
|
131 |
+
Product concept design (Product Design)
|
132 |
+
Marketing infographics (Infographic)
|
133 |
|
134 |
Education
|
135 |
|
136 |
+
Learning content organization (Concept Map, Mindmap)
|
137 |
Course explanatory materials (Synoptic Chart)
|
138 |
Research topic structuring (Concept Map)
|
139 |
+
Educational material visualization (Diagram)
|
140 |
|
141 |
Technical Documentation
|
142 |
|
143 |
System architecture diagramming
|
144 |
API flow explanation
|
145 |
Data processing pipeline visualization
|
146 |
+
UI/UX mockups (Mockup)
|
147 |
+
Technical flowcharts (Flowchart)
|
148 |
|
149 |
๐ง Advanced Configuration
|
150 |
Customizing JSON Structure
|
151 |
Each diagram type has its unique JSON structure:
|
152 |
Concept Map Example:
|
|
|
153 |
json{
|
154 |
"central_node": "Artificial Intelligence",
|
155 |
"nodes": [
|
|
|
161 |
}
|
162 |
]
|
163 |
}
|
|
|
164 |
Prompt Writing Tips
|
165 |
|
166 |
Be Specific: "AI classification system" โ "AI classification system including deep learning, machine learning, and natural language processing"
|
167 |
Specify Relationships: "A is part of B", "X leads to Y", etc.
|
168 |
Describe Hierarchical Structure: "A at the top level, with B and C below", etc.
|
169 |
+
For Design Types: Input in Korean is automatically translated to English
|
170 |
|
171 |
๐ก Model Performance
|
172 |
Gemma-3-R1984-27B Features
|
173 |
|
174 |
+
High Factuality
|
175 |
Structuring Ability: Accurately understands complex relationships and hierarchical structures
|
176 |
Multilingual Support: Ensures equal quality in both Korean and English
|
177 |
Context Understanding: Accurately grasps prompt intent
|
178 |
|
179 |
+
FLUX.1-schnell Features
|
180 |
+
|
181 |
+
High-Quality Images: Professional-level visual design generation
|
182 |
+
Fast Generation: Efficient inference for quick results
|
183 |
+
Diverse Styles: Support for sketches, wireframes, infographics, and various styles
|
184 |
+
|
185 |
๐ ๏ธ Tech Stack
|
186 |
|
187 |
Frontend: Gradio (with built-in MCP server)
|
188 |
+
AI Models:
|
189 |
+
|
190 |
+
Gemma-3-R1984-27B
|
191 |
+
FLUX.1-schnell (High-quality image generation)
|
192 |
+
|
193 |
+
|
194 |
Diagram Engine: Graphviz
|
195 |
Search Engine: Brave Search API
|
196 |
Protocol: MCP (Model Context Protocol)
|
197 |
Language: Python
|
198 |
|
|
|
199 |
<div align="center">
|
200 |
<p>Made with โค๏ธ by VIDraft Team</p>
|
201 |
<p><strong>Powered by VIDraftโก</strong></p>
|
202 |
<p><em>MCP-enabled for seamless AI integration</em></p>
|
203 |
+
</div>
|
|