File size: 1,740 Bytes
b651a1e bb445c8 b651a1e bb445c8 b651a1e 69730ef b651a1e |
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 |
---
language:
- en
- zh
tags:
- uml
- software-engineering
- requirements-modeling
- glm4
- code-generation
license: apache-2.0
---
# GLM4UML
GLM4UML is a fine-tuned version of GLM4-9B specifically designed for generating UML diagrams from natural language requirements. It supports class diagrams, use case diagrams, and sequence diagrams generation with high accuracy.
## Model Description
- **Model Type:** Fine-tuned GLM4-9B
- **Language(s):** English and Chinese
- **Task:** UML Diagram Generation
- **License:** Apache 2.0
### Training Data
The model was trained on a comprehensive dataset of software requirements and their corresponding UML diagrams, focusing on:
- Class diagrams
- Use case diagrams
- Sequence diagrams
## Performance and Limitations
### Metrics
The model achieves the following performance metrics:
- Precision: 78.68%
- Recall: 67.37%
- F1 Score: 72.59%
Compared to ChatGPT 4.0, AUG shows improved performance in:
- Recall rate
- Overall F1 score
### Limitations
While the model performs well, users should be aware that:
- The model's output may still require human verification
- Complex software architectures might need additional refinement
- Performance may vary based on the clarity of input requirements
## Training Procedure
The model was fine-tuned from GLM4-9B using:
- Specialized UML diagram dataset
- Requirements-to-diagram mapping
- Clarification dialogue patterns
## Additional Information
- **GitHub Repository:** https://github.com/XIAOLingQ/AUG
- **Demo Video:** https://youtu.be/kHbCPK6kOag
## Acknowledgements
This model was developed by researchers from Wuhan Textile University and Wuhan University. Special thanks to all contributors and the open-source community. |