Block and Subword-Scaling Floating-Point (BSFP) : An Efficient Non-Uniform Quantization For Low Precision Inference

Conference paper

允辰羅, 澤廣李 and 仁碩呂

Proceedings - International Conference on Learning Representations (ICLR) 2023

01/05/2023

Abstract

AI Accelerator, Block Floating Point

1 Record Views

Title: Block and Subword-Scaling Floating-Point (BSFP) : An Efficient Non-Uniform Quantization For Low Precision Inference
Creators - without role: 允辰羅 - 清大電機系
澤廣李 - 清大電機系
仁碩呂 - 清大電機系
Publication Details: Proceedings - International Conference on Learning Representations (ICLR) 2023
Identifiers: 9957770479006774
Academic Unit: Department of Electrical Engineering, College of Electrical Engineering and Computer Science, National Tsing Hua University
Language: English
Resource Type: Conference paper