When documenting AI applications that run local models or perform inference, always provide comprehensive infrastructure requirements including hardware specifications, platform-specific drivers, and setup instructions. This ensures users can successfully run AI models without encountering compatibility issues. Include the following details: - **Hardware requirements**: Specify RAM/VRAM needs for different model sizes (e.g., "8GB RAM/VRAM for 3B models, 16GB for 7B models") - **CPU/GPU architecture support**: List supported architectures (ARM, x86 for CPU; NVIDIA, AMD, Intel for GPU) - **Platform-specific drivers**: Include GPU driver requirements (NVIDIA drivers for Windows, CUDA Toolkit for Linux) - **Model size relationships**: Explain how hardware capacity relates to usable model sizes Example documentation structure: ```markdown ### Hardware Requirements - **RAM/VRAM**: 8GB minimum (3B models), 16GB recommended (7B models) - **CPU**: ARM, x86 architectures supported - **GPU**: NVIDIA (via llama.cpp), AMD and Intel support coming soon ### Platform Setup - **Windows**: Install NVIDIA drivers if GPU available - **Linux**: Install CUDA Toolkit if GPU available ``` This prevents user frustration and ensures AI applications can run as intended across different hardware configurations.