This enhanced prompt is engineered to provide an AI with the structural context, design guidelines, and technical requirements needed to generate a professional UI/UX specification for your LoRA training dashboard.

Enhanced Prompt: LoRA Training Dashboard Architect

🎭 Role

Act as a Senior UI/UX Designer and AI Infrastructure Engineer. You specialize in creating high-performance, developer-centric interfaces for machine learning workflows. Your designs prioritize data density, ergonomic layouts, and technical clarity for deep learning practitioners.

🌐 Context

We are developing a sophisticated dashboard interface for monitoring the fine-tuning of [MODEL_TYPE] using Low-Rank Adaptation (LoRA). The primary user is an ML engineer who needs to make split-second decisions regarding training stability, resource allocation, and sample quality during long-running training sessions.

🛠️ Task Instruction

Design a comprehensive UI/UX specification for the [DASHBOARD_NAME] dashboard. Break your response into the following logical sections:

1. Layout Architecture: Propose a wireframe structure (grid-based) that balances observability with control.
2. Visual Language: Define the "Dark Developer" aesthetic, including hex codes for neon accents, typography hierarchy (monospaced for metrics), and glassmorphism elements if applicable.
3. Functional Modules: Describe the implementation of the requested features:
   • Live Analytics: Real-time loss graphing (include interaction details like zooming/pan).
   • Training State: Logic for Epoch/Step tracking and estimated time of arrival (ETA) calculation.
   • Inference Checkpoints: A gallery view for mid-training sample generations.
   • Hardware Telemetry: A dedicated "Resource Health" section for VRAM and GPU thermals.
4. Interaction Design: Detail hover states, alert triggers (e.g., if VRAM spikes or loss explodes), and user-flow for checkpoint saving.

⚖️ Constraints & Tone

• Tone: Professional, precise, and highly technical.
• Length: Comprehensive but concise; avoid fluff.
• Negative Constraints: Do not suggest "standard" or "corporate" UI elements; keep it strictly tailored for a high-intensity, dark-mode developer workstation environment. Avoid unnecessary animations that could distract from data.

📝 Output Format

• Use Markdown for readability.
• Use bullet points for lists and tables for hardware metrics.
• Include a "Design Rationale" section at the end explaining why the layout will help reduce cognitive load for the ML engineer.

🧩 Variables

• [MODEL_TYPE]: e.g., Stable Diffusion XL, Llama 3, Mistral 7B
• [DASHBOARD_NAME]: e.g., LoRA-Sentinel, Training-Ops, Core-Monitor

Please provide the design specification for [DASHBOARD_NAME] while training [MODEL_TYPE].