Quantization noise in low bit quantization and iterative adaptation to quantization noise in quantizable neural networks

Standard

Quantization noise in low bit quantization and iterative adaptation to quantization noise in quantizable neural networks. / Chudakov, D.; Goncharenko, A.; Alyamkin, S. et al.

In: Journal of Physics: Conference Series, Vol. 2134, No. 1, 012004, 20.12.2021.

Research output: Contribution to journal › Conference article › peer-review

BibTeX

@article{9a1018eacb73487d8829c59e72f04459,

title = "Quantization noise in low bit quantization and iterative adaptation to quantization noise in quantizable neural networks",

abstract = "Quantization is one of the most popular and widely used methods of speeding up a neural network. At the moment, the standard is 8-bit uniform quantization. Nevertheless, the use of uniform low-bit quantization (4- and 6-bit quantization) has significant advantages in speed and resource requirements for inference. We present our quantization algorithm that offers advantages when using uniform low-bit quantization. It is faster than quantization-aware training from scratch and more accurate than methods aimed only at selecting thresholds and reducing noise from quantization. We also investigated quantization noise in neural networks for low-bit quantization and concluded that quantization noise is not always a good metric for quantization quality.",

author = "D. Chudakov and A. Goncharenko and S. Alyamkin and A. Densidov",

note = "Publisher Copyright: {\textcopyright} 2021 Institute of Physics Publishing. All rights reserved.; 8th International Young Scientists Conference on Information Technologies, Telecommunications and Control Systems, ITTCS 2021 ; Conference date: 16-12-2021 Through 17-12-2021",

year = "2021",

month = dec,

day = "20",

doi = "10.1088/1742-6596/2134/1/012004",

language = "English",

volume = "2134",

journal = "Journal of Physics: Conference Series",

issn = "1742-6588",

publisher = "IOP Publishing Ltd.",

number = "1",

}

RIS

TY - JOUR

T1 - Quantization noise in low bit quantization and iterative adaptation to quantization noise in quantizable neural networks

AU - Chudakov, D.

AU - Goncharenko, A.

AU - Alyamkin, S.

AU - Densidov, A.

PY - 2021/12/20

Y1 - 2021/12/20

N2 - Quantization is one of the most popular and widely used methods of speeding up a neural network. At the moment, the standard is 8-bit uniform quantization. Nevertheless, the use of uniform low-bit quantization (4- and 6-bit quantization) has significant advantages in speed and resource requirements for inference. We present our quantization algorithm that offers advantages when using uniform low-bit quantization. It is faster than quantization-aware training from scratch and more accurate than methods aimed only at selecting thresholds and reducing noise from quantization. We also investigated quantization noise in neural networks for low-bit quantization and concluded that quantization noise is not always a good metric for quantization quality.

AB - Quantization is one of the most popular and widely used methods of speeding up a neural network. At the moment, the standard is 8-bit uniform quantization. Nevertheless, the use of uniform low-bit quantization (4- and 6-bit quantization) has significant advantages in speed and resource requirements for inference. We present our quantization algorithm that offers advantages when using uniform low-bit quantization. It is faster than quantization-aware training from scratch and more accurate than methods aimed only at selecting thresholds and reducing noise from quantization. We also investigated quantization noise in neural networks for low-bit quantization and concluded that quantization noise is not always a good metric for quantization quality.

UR - http://www.scopus.com/inward/record.url?scp=85123640836&partnerID=8YFLogxK

U2 - 10.1088/1742-6596/2134/1/012004

DO - 10.1088/1742-6596/2134/1/012004

M3 - Conference article

AN - SCOPUS:85123640836

VL - 2134

JO - Journal of Physics: Conference Series

JF - Journal of Physics: Conference Series

SN - 1742-6588

IS - 1

M1 - 012004

T2 - 8th International Young Scientists Conference on Information Technologies, Telecommunications and Control Systems, ITTCS 2021

Y2 - 16 December 2021 through 17 December 2021

ER -

ID: 35377964