A compact, efficient embedding specialist that converts text into dense vector representations. It operates with a lightweight footprint, making it well-suited for on-device inference via the ExecuTorch runtime. Its small size means it trades some representational depth for speed and portability.