id	author	title	date	pages	extension	mime	words	sentence	flesch	summary	cache	txt
ajst-12694	Wang, Hanyun	Investigating Lightweight Transformer Models for Defect Detection	2023	7	.pdf	application/pdf	4746	279	47	Second, detection models tend to be large, demanding high-performance hardware, incurring high costs, and operating slowly, making them unsuitable for real-time applications. Pruning-based model lightweighting method Pruning is an effective method to reduce the enormous inference cost of Transformer models.	cache/ajst-12694.pdf	txt/ajst-12694.txt