{"title":"镶嵌:资源高效llm的复合投影剪枝","authors":"Bailey J. Eccles , Leon Wong , Blesson Varghese","doi":"10.1016/j.future.2025.108056","DOIUrl":null,"url":null,"abstract":"<div><div>Extensive compute and memory requirements limit the deployment of large language models (LLMs) on any hardware. Compression methods, such as pruning, can reduce model size, which in turn reduces resource requirements. State-of-the-art pruning is based on coarse-grained methods. They are time-consuming and inherently remove critical model parameters, adversely impacting the quality of the pruned model. This paper introduces projection pruning, a novel fine-grained method for pruning LLMs. In addition, LLM projection pruning is enhanced by a new approach we refer to as composite projection pruning — the synergistic combination of unstructured pruning that retains accuracy and structured pruning that reduces model size. We develop <span><span>Mosaic</span></span>, a novel system to create and deploy pruned LLMs using composite projection pruning. <span><span>Mosaic</span></span> is evaluated using a range of performance and quality metrics on multiple hardware platforms, LLMs, and datasets. <span><span>Mosaic</span></span> is 7.19<span><math><mo>×</mo></math></span> faster in producing models than existing approaches. <span><span>Mosaic</span></span> models achieve up to 84.2% lower perplexity and 31.4% higher accuracy than models obtained from coarse-grained pruning. Up to 67% faster inference and 68% lower GPU memory use is noted for <span><span>Mosaic</span></span> models.</div></div>","PeriodicalId":55132,"journal":{"name":"Future Generation Computer Systems-The International Journal of Escience","volume":"175 ","pages":"Article 108056"},"PeriodicalIF":6.2000,"publicationDate":"2025-08-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Mosaic: Composite projection pruning for resource-efficient LLMs\",\"authors\":\"Bailey J. Eccles , Leon Wong , Blesson Varghese\",\"doi\":\"10.1016/j.future.2025.108056\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"<div><div>Extensive compute and memory requirements limit the deployment of large language models (LLMs) on any hardware. Compression methods, such as pruning, can reduce model size, which in turn reduces resource requirements. State-of-the-art pruning is based on coarse-grained methods. They are time-consuming and inherently remove critical model parameters, adversely impacting the quality of the pruned model. This paper introduces projection pruning, a novel fine-grained method for pruning LLMs. In addition, LLM projection pruning is enhanced by a new approach we refer to as composite projection pruning — the synergistic combination of unstructured pruning that retains accuracy and structured pruning that reduces model size. We develop <span><span>Mosaic</span></span>, a novel system to create and deploy pruned LLMs using composite projection pruning. <span><span>Mosaic</span></span> is evaluated using a range of performance and quality metrics on multiple hardware platforms, LLMs, and datasets. <span><span>Mosaic</span></span> is 7.19<span><math><mo>×</mo></math></span> faster in producing models than existing approaches. <span><span>Mosaic</span></span> models achieve up to 84.2% lower perplexity and 31.4% higher accuracy than models obtained from coarse-grained pruning. Up to 67% faster inference and 68% lower GPU memory use is noted for <span><span>Mosaic</span></span> models.</div></div>\",\"PeriodicalId\":55132,\"journal\":{\"name\":\"Future Generation Computer Systems-The International Journal of Escience\",\"volume\":\"175 \",\"pages\":\"Article 108056\"},\"PeriodicalIF\":6.2000,\"publicationDate\":\"2025-08-11\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Future Generation Computer Systems-The International Journal of Escience\",\"FirstCategoryId\":\"94\",\"ListUrlMain\":\"https://www.sciencedirect.com/science/article/pii/S0167739X25003516\",\"RegionNum\":2,\"RegionCategory\":\"计算机科学\",\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q1\",\"JCRName\":\"COMPUTER SCIENCE, THEORY & METHODS\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Future Generation Computer Systems-The International Journal of Escience","FirstCategoryId":"94","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S0167739X25003516","RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"COMPUTER SCIENCE, THEORY & METHODS","Score":null,"Total":0}
Mosaic: Composite projection pruning for resource-efficient LLMs
Extensive compute and memory requirements limit the deployment of large language models (LLMs) on any hardware. Compression methods, such as pruning, can reduce model size, which in turn reduces resource requirements. State-of-the-art pruning is based on coarse-grained methods. They are time-consuming and inherently remove critical model parameters, adversely impacting the quality of the pruned model. This paper introduces projection pruning, a novel fine-grained method for pruning LLMs. In addition, LLM projection pruning is enhanced by a new approach we refer to as composite projection pruning — the synergistic combination of unstructured pruning that retains accuracy and structured pruning that reduces model size. We develop Mosaic, a novel system to create and deploy pruned LLMs using composite projection pruning. Mosaic is evaluated using a range of performance and quality metrics on multiple hardware platforms, LLMs, and datasets. Mosaic is 7.19 faster in producing models than existing approaches. Mosaic models achieve up to 84.2% lower perplexity and 31.4% higher accuracy than models obtained from coarse-grained pruning. Up to 67% faster inference and 68% lower GPU memory use is noted for Mosaic models.
期刊介绍:
Computing infrastructures and systems are constantly evolving, resulting in increasingly complex and collaborative scientific applications. To cope with these advancements, there is a growing need for collaborative tools that can effectively map, control, and execute these applications.
Furthermore, with the explosion of Big Data, there is a requirement for innovative methods and infrastructures to collect, analyze, and derive meaningful insights from the vast amount of data generated. This necessitates the integration of computational and storage capabilities, databases, sensors, and human collaboration.
Future Generation Computer Systems aims to pioneer advancements in distributed systems, collaborative environments, high-performance computing, and Big Data analytics. It strives to stay at the forefront of developments in grids, clouds, and the Internet of Things (IoT) to effectively address the challenges posed by these wide-area, fully distributed sensing and computing systems.