Adaptive Feature-based Low-Rank Compression of Large Language Models via Bayesian Optimization
This work addresses efficiency challenges for users of large language models, but it is incremental as it builds on existing low-rank compression methods.
The paper tackles the computational burden of large language models by proposing a low-rank compression method that uses pooled covariance matrices and Bayesian optimization to allocate low-rank dimensions, achieving better performance than existing techniques at the same compression ratio on LLaMA-2 models.
In recent years, large language models (LLMs) have driven advances in natural language processing. Still, their growing scale has increased the computational burden, necessitating a balance between efficiency and performance. Low-rank compression, a promising technique, reduces non-essential parameters by decomposing weight matrices into products of two low-rank matrices. Yet, its application in LLMs has not been extensively studied. The key to low-rank compression lies in low-rank factorization and low-rank dimensions allocation. To address the challenges of low-rank compression in LLMs, we conduct empirical research on the low-rank characteristics of large models. We propose a low-rank compression method suitable for LLMs. This approach involves precise estimation of feature distributions through pooled covariance matrices and a Bayesian optimization strategy for allocating low-rank dimensions. Experiments on the LLaMA-2 models demonstrate that our method outperforms existing strong structured pruning and low-rank compression techniques in maintaining model performance at the same compression ratio.