LLM Compression: How Far Can We Go in Balancing Size and Performance?

TL;DR

This paper evaluates 4-bit quantization techniques, GSQ and GPTQ, on various LLMs to analyze the trade-offs between model size reduction and task performance across multiple NLP benchmarks.

Contribution

It provides a comprehensive benchmark of GSQ and GPTQ quantization methods on different LLMs, assessing their impact on accuracy and efficiency for real-world NLP tasks.

Findings

Quantization reduces model size and inference latency.

Trade-offs exist between compression level and task accuracy.

GSQ and GPTQ have different strengths depending on model size.

Abstract

Quantization is an essential and popular technique for improving the accessibility of large language models (LLMs) by reducing memory usage and computational costs while maintaining performance. In this study, we apply 4-bit Group Scaling Quantization (GSQ) and Generative Pretrained Transformer Quantization (GPTQ) to LLaMA 1B, Qwen 0.5B, and PHI 1.5B, evaluating their impact across multiple NLP tasks. We benchmark these models on MS MARCO (Information Retrieval), BoolQ (Boolean Question Answering), and GSM8K (Mathematical Reasoning) datasets, assessing both accuracy and efficiency across various tasks. The study measures the trade-offs between model compression and task performance, analyzing key evaluation metrics, namely accuracy, inference latency, and throughput (total output tokens generated per second), providing insights into the suitability of low-bit quantization for real-world deployment. Using the results, users can then make suitable decisions based on the specifications that need to be met. We discuss the pros and cons of GSQ and GPTQ techniques on models of different sizes, which also serve as a benchmark for future experiments.