Hugging Face Overview
Model Overview
sequelbox/Qwen3-14B-UML-Generator is a 14 billion parameter model based on the Qwen3 architecture, developed by sequelbox. It is a specialized, experimental reasoning model designed to generate Unified Modeling Language (UML) diagrams from natural language descriptions. Unlike general-purpose chat models, this finetuned variant focuses exclusively on producing structured XMI 2.5.1 code for UML diagrams.
Key Capabilities
- UML Diagram Generation: Translates situational analysis and reasoning into XMI 2.5.1 code for various UML diagram types.
- Multi-step Reasoning: Employs an internal multi-step reasoning process to identify diagram structure before generating the final output.
- Specialized Finetuning: Trained on sequelbox's custom UML creation dataset, leveraging data generated with DeepSeek-V3.2.
- Broad Application Domains: Capable of analyzing and generating diagrams for software architecture, development, business processes, systems engineering, data modeling, microservices, and reverse engineering.
- Optimized Output Format: Provides a clear, readable XMI output in a dedicated "UML Generator Format" ready for direct loading into UML tools.
Good For
- Developers and architects needing to quickly generate UML diagrams from textual descriptions.
- Integrating automated UML generation into development workflows.
- Analyzing complex systems and processes to visualize their structure and behavior through diagrams.
- Use cases requiring a specialized code-reasoning assistant for diagrammatic representation.