A Token is Worth over 1,000 Tokens: Efficient Knowledge Distillation through Low-Rank Clone

TL;DR

This paper introduces Low-Rank Clone (LRC), a novel pre-training method that significantly improves the efficiency of training small language models by effectively transferring knowledge from large teachers using low-rank projections and activation cloning.

Contribution

LRC combines low-rank weight compression and activation alignment to enhance knowledge distillation, reducing training data needs and computational costs for small language models.

Findings

LRC achieves comparable or better performance than state-of-the-art models.

LRC reduces training tokens from trillions to 20 billion.

LRC improves training efficiency by over 1,000 times.

Abstract

Training high-performing Small Language Models (SLMs) remains costly, even with knowledge distillation and pruning from larger teacher models. Existing work often faces three key challenges: (1) information loss from hard pruning, (2) inefficient alignment of representations, and (3) underutilization of informative activations, particularly from Feed-Forward Networks (FFNs). To address these challenges, we introduce Low-Rank Clone (LRC), an efficient pre-training method that constructs SLMs aspiring to behavioral equivalence with strong teacher models. LRC trains a set of low-rank projection matrices that jointly enable soft pruning by compressing teacher weights, and activation clone by aligning student activations, including FFN signals, with those of the teacher. This unified design maximizes knowledge transfer while removing the need for explicit alignment modules. Extensive experiments with open-source teachers (e.g., Llama-3.2-3B-Instruct, Qwen2.5-3B/7B-Instruct) show that LRC matches or surpasses state-of-the-art models trained on trillions of tokens--while using only 20B tokens, achieving over 1,000x training efficiency. Our codes and model checkpoints are available at https://github.com/CURRENTF/LowRankClone and https://huggingface.co/collections/JitaiHao/low-rank-clone-lrc-6828389e96a93f1d4219dfaf.