CBQ: Cross-Block Quantization for Large Language Models
This addresses the challenge of efficient compression for large language models, offering a practical solution for deployment with improved accuracy and speed, though it is incremental as it builds on existing PTQ methods.
The paper tackles the problem of severe performance degradation in low-bit post-training quantization for large language models by proposing CBQ, a cross-block reconstruction-based method that establishes long-range dependencies across blocks; it achieves superior low-bit quantization results, such as quantizing a 4-bit LLAMA1-65B model in 4.3 hours on a single GPU.
Post-training quantization (PTQ) has played a key role in compressing large language models (LLMs) with ultra-low costs. However, existing PTQ methods only focus on handling the outliers within one layer or one block, which ignores the dependency of blocks and leads to severe performance degradation in low-bit settings. In this paper, we propose CBQ, a cross-block reconstruction-based PTQ method for LLMs. CBQ employs a cross-block dependency using a homologous reconstruction scheme, establishing long-range dependencies across multiple blocks to minimize error accumulation. Furthermore, CBQ incorporates a coarse-to-fine preprocessing (CFP) strategy for suppressing weight and activation outliers, coupled with an adaptive LoRA-Rounding technique for precise weight quantization. These innovations enable CBQ to not only handle extreme outliers effectively but also improve overall quantization accuracy. Extensive experiments show that CBQ achieves superior low-bit quantization (W4A4, W4A8, W2A16) and outperforms existing state-of-the-art methods across various LLMs and datasets. Notably, CBQ quantizes the 4-bit LLAMA1-65B model within only 4.3 hours on a single GPU, achieving a commendable tradeoff between performance and quantization efficiency.