LLM-FP4: 4-Bit Floating-Point Quantized Transformers.

Shih-Yang Liu Zechun Liu Xijie Huang Pingcheng Dong Kwang-Ting Cheng

Published in: CoRR (2023)

Keyphrases

floating point
fixed point
square root
significant bit
sparse matrices
dct coefficients
fast fourier transform
instruction set
pairwise
general purpose
floating point arithmetic