ZeroQuant-FP: A Leap Forward in LLMs Post-Training W4A8 Quantization Using Floating-Point Formats.

Xiaoxia Wu Zhewei Yao Yuxiong He

Published in: CoRR (2023)

Keyphrases

floating point
floating point arithmetic
fixed point
square root
training set
fast fourier transform
sparse matrices
metadata
signal processing
image processing
interval arithmetic