UniForm: A Reuse Attention Mechanism Optimized for Efficient Vision Transformers on Edge Devices

Bibliografski detalji

Izdano u:	arXiv.org (Dec 3, 2024), p. n/a
Glavni autor:	Seul-Ki Yeom
Daljnji autori:	Tae-Ho, Kim
Izdano:	Cornell University Library, arXiv.org
Teme:	Accuracy Attention Memory tasks Memory devices Flash memory (computers) Platforms Real time Task complexity Inference
Online pristup:	Citation/Abstract Full text outside of ProQuest
Oznake:	Dodaj oznaku Bez oznaka, Budi prvi tko označuje ovaj zapis!