Model Selection and Prompting

This unit examines the tradeoffs between large, compact, and domain-specific language models and how to leverage them effectively in Python-based solutions. Students compare real-world examples, practice selecting appropriate model types for task constraints, explore prompting strategies with general-purpose models, and receive an introduction to tuning approaches to prepare for more advanced model customization.

Learning Objectives

• Analyze tradeoffs among large, compact, and domain-specific models with respect to performance, latency, cost, privacy, and deployment constraints
• Apply selection criteria to choose between general-purpose, compact, or domain-specific models for given Python solution scenarios
• Evaluate real-world examples of successful smaller and domain-specific models to identify patterns of performance, efficiency, and appropriate use cases
• Demonstrate effective prompting techniques for general-purpose models and outline when tuning strategies should be pursued