Outlier-Safe Pre-Training for Robust 4-Bit Quantization of Large Language Models

TL;DR

This paper introduces Outlier-Safe Pre-Training (OSP), a novel training approach that prevents activation outliers in large language models, enabling more effective 4-bit quantization and efficient deployment.

Contribution

OSP combines three innovations to proactively prevent outliers during training, allowing large language models to be quantized efficiently without post-hoc mitigation.

Findings

First production-scale outlier-free LLM trained on 1 trillion tokens.

Achieves 35.7 average score on 10 benchmarks under 4-bit quantization.

Near-zero excess kurtosis (0.04) indicating drastically reduced outliers.

Abstract

Extreme activation outliers in Large Language Models (LLMs) critically degrade quantization performance, hindering efficient on-device deployment. While channel-wise operations and adaptive gradient scaling are recognized causes, practical mitigation remains challenging. We introduce Outlier-Safe Pre-Training (OSP), a practical guideline that proactively prevents outlier formation rather than relying on post-hoc mitigation. OSP combines three key innovations: (1) the Muon optimizer, eliminating privileged bases while maintaining training efficiency; (2) Single-Scale RMSNorm, preventing channel-wise amplification; and (3) a learnable embedding projection, redistributing activation magnitudes originating from embedding matrices. We validate OSP by training a 1.4B-parameter model on 1 trillion tokens, which is the first production-scale LLM trained without such outliers. Under aggressive 4-bit quantization, our OSP model achieves a 35.7 average score across 10 benchmarks (compared to 26.5 for an Adam-trained model), with only a 2% training overhead. Remarkably, OSP models exhibit near-zero excess kurtosis (0.04) compared to extreme values (1818.56) in standard models, fundamentally altering LLM quantization behavior. Our work demonstrates that outliers are not inherent to LLMs but are consequences of training strategies, paving the way for more efficient LLM deployment. The source code and pretrained checkpoints are available at https://github.com/dmis-lab/Outlier-Safe-Pre-Training.