Coding and Gaming Design - Exploring TensorFlow and Hugging Face

Exploring TensorFlow and Hugging Face gives students a window into the pre-trained AI systems powering today's most advanced applications—without requiring them to build those systems from scratch. Students interact directly with real AI models to investigate how they process inputs, generate outputs, and reflect the patterns embedded in their training data.

Students begin with an introduction to what pre-trained models are and why they matter: rather than training a model from zero (which requires enormous data and compute), developers and researchers use and fine-tune existing models that have already learned from massive datasets. TensorFlow Hub and Hugging Face are two of the most widely used platforms for accessing these models.

Through structured exploration, students run models, test inputs, observe outputs, and ask questions about what's happening inside the system. They investigate text classification, image recognition, and language generation models—developing intuitions about what different architectures do well and where their limits lie.

The lesson puts significant emphasis on what training data reveals: students examine how models reflect the patterns, assumptions, and gaps in the data they were trained on—and what that means for the communities and contexts where these models are deployed.

Students leave with direct experience with professional-grade AI tools and a more sophisticated mental model of what AI systems actually are and how they work.

ISTE Standards 1 and 5 aligned.