# US 2021/0081783 A1 · Information processing apparatus, method of processing information, and non-transitory computer-readable storage medium for storing information processing program

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2021/0081783 A1
- Document ID: US20210081783A1
- Status: Published application
- Kind: A1
- Application number: 16/930,356
- Filed: 2020-07-16
- Published: 2021-03-18
- Earliest claimed: 2019-09-13
- Assignees: Fujitsu Limited
- Applicants: Fujitsu Limited
- Inventors: Yasufumi Sakai; Sosaku Moriki
- Web page: https://patentgrep.com/patent/US20210081783A1

## CPC

- H03M 7/3059: {Digital compression and data reduction techniques where the original information is represented by a subset or similar information, e.g. lossy compression}
- G06N 3/04: Architecture, e.g. interconnection topology
- G06N 3/045: Combinations of networks
- G06N 3/0464: Convolutional networks [CNN, ConvNet]
- G06N 3/0495: Quantised networks; Sparse networks; Compressed networks
- G06N 3/063: using electronic means
- G06N 3/08: Learning methods
- G06N 3/084: Backpropagation, e.g. using gradient descent
- G06N 3/09: Supervised learning
- H03M 7/70: {Type of the data to be coded, other than image and sound}
- G06F 18/21: Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- H03M: CODING; DECODING; CODE CONVERSION IN GENERAL (using fluidic means F15C4/00; optical analogue/digital converters G02F7/00; coding, decoding or code conversion, specially adapted for particular applications, see the relevant subclasses, e.g. G01D, G01R, G06F, G06T, G09G, G10L, G11B, G11C, H04B, H04L, H04M, H04N; ciphering or deciphering for cryptography or other purposes involving the need for secrecy G09C)
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)

## Abstract

An information processing apparatus includes: a memory configured to store program instructions to perform quantization on quantization target data; and a processor configured to execute the program instructions stored in the memory, the program instructions including: obtaining a distribution of appearance frequencies of a plurality of variable elements included in the quantization target data; and aligning a most significant bit position of a quantization position to a variable element smaller than a variable element of a maximum value among the plurality of variable elements based on the distribution of the appearance frequencies of the plurality of variable elements.

## Claims

### Claim 1

An information processing apparatus comprising:
a memory configured to store program instructions to perform quantization on quantization target data; and
a processor configured to execute the program instructions stored in the memory, the program instructions including:
obtaining a distribution of appearance frequencies of a plurality of variable elements included in the quantization target data; and
aligning a most significant bit position of a quantization position to a variable element smaller than a variable element of a maximum value among the plurality of variable elements based on the distribution of the appearance frequencies of the plurality of variable elements.

### Claim 2 (depends on 1)

The information processing apparatus according to claim 1,
wherein the aligning is performed such that a value outside a quantization range becomes the most significant bit position after the quantization.

### Claim 3 (depends on 1)

The information processing apparatus according claim 1, wherein
the quantization target data is a gradient used for updating a weight of a neural network.

### Claim 4 (depends on 1)

The information processing apparatus according to claim 1, wherein,
in response to detecting that a learning rate of a neural network is smaller than a predetermined threshold, the aligning is performed.

### Claim 5

A computer implemented-method of processing information comprising:
when performing quantization on quantization target data
obtaining, by a computer, a distribution of appearance frequencies of a plurality of variable elements included in the quantization target data; and
aligning, by the computer, a most significant bit position of a quantization position to a variable element smaller than a variable element of a maximum value among the plurality of variable elements based on the distribution of the appearance frequencies of the plurality of variable elements.

### Claim 6

A non-transitory computer-readable storage medium for storing an information processing program which causes a processor to perform processing, the processing comprising:
when performing quantization on quantization target data,
obtaining a distribution of appearance frequencies of a plurality of variable elements included in the quantization target data, and
aligning a most significant bit position of a quantization position to a variable element smaller than a variable element of a maximum value among the plurality of variable elements based on the distribution of the appearance frequencies of the plurality of variable elements.
