What is SLM (Small Language Model)?
TL;DR
A lightweight language model with fewer parameters. Excels in efficiency and cost-effectiveness.
SLM (Small Language Model): Definition & Explanation
An SLM (Small Language Model) is a relatively lightweight language model with hundreds of millions to a few billion parameters. While LLMs (Large Language Models) have hundreds of billions of parameters, SLMs are designed to run on limited computational resources and are well-suited for edge devices and mobile phones. Notable examples include Microsoft's Phi-3, Google's Gemma, and Meta's LLaMA 3.2 (1B/3B). Through knowledge distillation, quantization, and efficient architecture design, SLMs achieve impressive performance relative to their size. When specialized for specific tasks, they can often match LLM-level accuracy, making them increasingly popular among cost-conscious enterprises. SLMs shine in scenarios where privacy is paramount, low latency is required, or internet connectivity is limited — use cases where LLMs are less practical. The evolution of SLMs is accelerating the democratization of AI.