A minimal, lightweight text model built for accessibility and experimentation. Its small footprint makes it easy to run locally without significant hardware requirements, though its capabilities are constrained by its compact size. Think of it as a starting point for understanding decoder architectures rather than a production workhorse.