Small language models are typically trained or fine-tuned for specific enterprise tasks rather than open-ended conversations ...