Model Overview

Tesslate/UIGEN-T2-7B is a 7.6 billion parameter model, fine-tuned from the Qwen2.5-Coder-7B-Instruct base using PEFT/LoRA, specifically engineered for UI code generation. This model specializes in producing HTML and Tailwind CSS for web interfaces, leveraging a substantial training dataset of 50,000 UI examples.

Key Capabilities

• High-Quality UI Code Generation: Produces functional and semantic HTML with utility-first Tailwind CSS.
• Massive Training Dataset: Trained on 50,000 diverse UI examples, enhancing component understanding and stylistic range.
• Innovative UI-Based Reasoning: Incorporates detailed reasoning traces from a specialized "teacher" model, ensuring outputs consider usability, layout, and aesthetics.
• PEFT/LoRA Trained: Efficiently fine-tuned with published LoRA checkpoints for transparency.
• Improved Chat Interaction: Features a streamlined prompt flow for more natural interaction.

Recommended Use Cases

• Rapid UI Prototyping: Quickly generate HTML/Tailwind code snippets from descriptions.
• Component Generation: Create standard and custom UI components (buttons, cards, forms).
• Frontend Development Assistance: Accelerate development by generating baseline component structures.
• Design-to-Code Exploration: Bridge design concepts with initial code implementation.

Limitations

• Primarily generates HTML and Tailwind CSS; Bootstrap support is planned.
• Focuses on structure and styling; complex JavaScript logic requires manual implementation.
• May need further fine-tuning for strict adherence to highly specific design systems.