[HN Gopher] Outlier Weighed Layerwise Sparsity: A Missing Secret...
___________________________________________________________________
Outlier Weighed Layerwise Sparsity: A Missing Secret Sauce for
Pruning LLMs
Author : amilios
Score : 4 points
Date : 2023-10-10 20:04 UTC (2 hours ago)
(HTM) web link (arxiv.org)
(TXT) w3m dump (arxiv.org)
| amilios wrote:
| Paper abstract: Large Language Models (LLMs), renowned for their
| remarkable performance across diverse domains, present a
| challenge due to their colossal model size when it comes to
| practical deployment. In response to this challenge, efforts have
| been directed toward the application of traditional network
| pruning techniques to LLMs, uncovering a massive number of
| parameters can be pruned in one-shot without hurting performance.
| Building upon insights gained from pre-LLM models, particularly
| BERT-level language models, prevailing LLM pruning strategies
| have consistently adhered to the practice of uniformly pruning
| all layers at equivalent sparsity levels, resulting in robust
| performance. However, this observation stands in contrast to the
| prevailing trends observed in the field of vision models, where
| non-uniform layerwise sparsity typically yields substantially
| improved results. To elucidate the underlying reasons for this
| disparity, we conduct a comprehensive analysis of the
| distribution of token features within LLMs. In doing so, we
| discover a strong correlation with the emergence of outliers,
| defined as features exhibiting significantly greater magnitudes
| compared to their counterparts in feature dimen- sions. Inspired
| by this finding, we introduce a novel LLM pruning methodology
| that incorporates a tailored set of non-uniform layerwise
| sparsity ratios specif- ically designed for LLM pruning, termed
| as Outlier Weighed Layerwise sparsity (OWL). The sparsity ratio
| of OWL is directly proportional to the outlier ratio observed
| within each layer, facilitating a more effective alignment
| between layer- wise weight sparsity and outlier ratios. Our
| empirical evaluation, conducted across the LLaMA-V1 family and
| OPT, spanning various benchmarks, demonstrates the distinct
| advantages offered by OWL over previous methods. For instance,
| our approach exhibits a remarkable performance gain, surpassing
| the state-of-the-art Wanda and SparseGPT by 61.22 and 6.80
| perplexity at a high sparsity level of 70%, respectively. Codes
| are available at https://github.com/luuyin/OWL.
___________________________________________________________________
(page generated 2023-10-10 23:02 UTC)