Výsledky vyhledávání

Akademický článek

Experimental study on the freezing process of water droplets for ice air jet technology

Autor: Hu Jingru, Li Jingbin, Huang Zhongwei, Cheng Kang, Xia Haojun

Publikováno v: Scientific Reports, Vol 14, Iss 1, Pp 1-10 (2024)

Abstract Ice air jet technology is one non-destructive, residue-free and environmentally friendly machining process. It is an efficient method to form ice particles by atomizing water droplets in ultra-low temperature environment. The freezing charac

Externí odkaz: https://doaj.org/article/d74494d47b724fc69a1f5270d3b077eb

Zobrazit plný text záznamu

Plný text ve formátu HTML

Report

FP6-LLM: Efficiently Serving Large Language Models Through FP6-Centric Algorithm-System Co-Design

Autor: Xia, Haojun, Zheng, Zhen, Wu, Xiaoxia, Chen, Shiyang, Yao, Zhewei, Youn, Stephen, Bakhtiari, Arash, Wyatt, Michael, Zhuang, Donglin, Zhou, Zhongzhu, Ruwase, Olatunji, He, Yuxiong, Song, Shuaiwen Leon

Six-bit quantization (FP6) can effectively reduce the size of large language models (LLMs) and preserve the model quality consistently across varied applications. However, existing systems do not provide Tensor Core support for FP6 quantization and s

Externí odkaz: http://arxiv.org/abs/2401.14112

Zobrazit plný text záznamu

Report

ZeroQuant(4+2): Redefining LLMs Quantization with a New FP6-Centric Strategy for Diverse Generative Tasks

Autor: Wu, Xiaoxia, Xia, Haojun, Youn, Stephen, Zheng, Zhen, Chen, Shiyang, Bakhtiari, Arash, Wyatt, Michael, Aminabadi, Reza Yazdani, He, Yuxiong, Ruwase, Olatunji, Song, Leon, Yao, Zhewei

This study examines 4-bit quantization methods like GPTQ in large language models (LLMs), highlighting GPTQ's overfitting and limited enhancement in Zero-Shot tasks. While prior works merely focusing on zero-shot measurement, we extend task scope to

Externí odkaz: http://arxiv.org/abs/2312.08583

Zobrazit plný text záznamu

Report

Flash-LLM: Enabling Cost-Effective and Highly-Efficient Large Generative Model Inference with Unstructured Sparsity

Autor: Xia, Haojun, Zheng, Zhen, Li, Yuchao, Zhuang, Donglin, Zhou, Zhongzhu, Qiu, Xiafei, Li, Yong, Lin, Wei, Song, Shuaiwen Leon

With the fast growth of parameter size, it becomes increasingly challenging to deploy large generative models as they typically require large GPU memory consumption and massive computation. Unstructured model pruning has been a common approach to red

Externí odkaz: http://arxiv.org/abs/2309.10285

Zobrazit plný text záznamu

Akademický článek

Experimental study on the Mode l fracture toughness of frozen silty clay incorporating Digital image correlation

Autor: Hua, Weihang, Zhang, Huimei, Wang, Yongtao, Xia, Haojun

Publikováno v: In Engineering Fracture Mechanics 25 November 2024 311

Zobrazit plný text záznamu

Report

Shift-BNN: Highly-Efficient Probabilistic Bayesian Neural Network Training via Memory-Friendly Pattern Retrieving

Autor: Wan, Qiyu, Xia, Haojun, Zhang, Xingyao, Wang, Lening, Song, Shuaiwen Leon, Fu, Xin

Bayesian Neural Networks (BNNs) that possess a property of uncertainty estimation have been increasingly adopted in a wide range of safety-critical AI applications which demand reliable and robust decision making, e.g., self-driving, rescue robots, m

Externí odkaz: http://arxiv.org/abs/2110.03553

Zobrazit plný text záznamu