Norm Tweaking: High-performance Low-bit Quantization of Large Language Models
This addresses the challenge of deploying large LLMs efficiently by enabling high-performance low-bit quantization, making it more practical for real-world applications, though it is incremental as it builds on existing PTQ methods.
The paper tackles the problem of severe performance degradation in low-bit quantization of large language models (LLMs) by introducing norm tweaking, a plugin technique that rectifies quantized activation distributions to match float counterparts, achieving the same accuracy as float models at 2-bit quantization on GLM-130B and OPT-66B.
As the size of large language models (LLMs) continues to grow, model compression without sacrificing accuracy has become a crucial challenge for deployment. While some quantization methods, such as GPTQ, have made progress in achieving acceptable 4-bit weight-only quantization, attempts at lower-bit quantization often result in severe performance degradation. In this paper, we introduce a technique called norm tweaking, which can be used as a plugin in current PTQ methods to achieve high precision while being cost-efficient. Our approach is inspired by the observation that rectifying the quantized activation distribution to match its float counterpart can readily restore accuracy for LLMs. To achieve this, we carefully design a tweaking strategy that includes calibration data generation and channel-wise distance constraint to update the weights of normalization layers for better generalization. We conduct extensive experiments on various datasets using several open-sourced LLMs. Our method demonstrates significant improvements in both weight-only quantization and joint quantization of weights and activations, surpassing existing PTQ methods. On GLM-130B and OPT-66B, our method even achieves the same level of accuracy at 2-bit quantization as their float ones. Our simple and effective approach makes it more practical for real-world applications.