Rethinking Channel Dimensions to Isolate Outliers for Low-bit Weight Quantization of Large Language Models

TL;DR

This paper introduces a novel quantization scheme called AdaDim that isolates outliers by rethinking channel dimensions, significantly improving low-bit weight quantization for large language models in resource-constrained environments.

Contribution

It proposes per-IC quantization and AdaDim framework, enhancing outlier handling and weight sensitivity adaptation for more effective low-bit quantization of LLMs.

Findings

Achieves up to +4.7% on MMLU benchmark

Improves performance on HumanEval by up to +10%

Demonstrates effectiveness across various language models

Abstract

Large Language Models (LLMs) have recently demonstrated remarkable success across various tasks. However, efficiently serving LLMs has been a challenge due to the large memory bottleneck, specifically in small batch inference settings (e.g. mobile devices). Weight-only quantization can be a promising approach, but sub-4 bit quantization remains a challenge due to large-magnitude activation outliers. To mitigate the undesirable outlier effect, we first propose per-IC quantization, a simple yet effective method that creates quantization groups within each input channel (IC) rather than the conventional per-output-channel (per-OC). Our method is motivated by the observation that activation outliers affect the input dimension of the weight matrix, so similarly grouping the weights in the IC direction can isolate outliers within a group. We also find that activation outliers do not dictate quantization difficulty, and inherent weight sensitivities also exist. With per-IC quantization as a new outlier-friendly scheme, we propose Adaptive Dimensions (AdaDim), a versatile quantization framework that can adapt to various weight sensitivity patterns. We demonstrate the effectiveness of AdaDim by augmenting prior methods such as Round-To-Nearest and GPTQ, showing significant improvements across various language modeling benchmarks for both base (up to +4.7% on MMLU) and instruction-tuned (up to +10% on HumanEval) LLMs. Code is available at https://github.com/johnheo/adadim-llm