EPO Patent Application: AI Model Training System

ChangeBridge: EPO Bulletin - AI & Computing (G06N)

Published March 18th, 2026

Detected March 23rd, 2026

Summary

The European Patent Office published patent application EP4711939A1 concerning systems and methods for training artificial intelligence models and storing checkpoint files. The application, filed by Alipay (Hangzhou) Information Technology Co., Ltd., details a system with a model training module and a cache module designed to efficiently manage datasets and AI model checkpoints.

View original document View source feed page

What changed

This document is a European Patent Office (EPO) publication (EP4711939A1) detailing a system and method for training artificial intelligence models and storing checkpoint files. The invention, assigned to Alipay (Hangzhou) Information Technology Co., Ltd., describes a system comprising a model training module and a first cache module. The training module utilizes a GPU chip to process datasets read from the cache module and generates checkpoint files, which are then sent back to the cache module. The cache module is designed to identify data types and store datasets locally from a buffer, while checkpoint files are directly stored on a local hard disk.

This patent application does not impose any new regulatory obligations or compliance requirements on regulated entities. It represents a disclosure of an invention related to AI infrastructure and data management. Compliance officers should note this as a development in AI technology but no immediate action is required.

Source document (simplified)

← EPO Patent Bulletin

SYSTEMS FOR TRAINING ARTIFICIAL INTELLIGENCE MODEL AND CHECKPOINT FILE STORAGE METHODS

Publication EP4711939A1 Kind: A1 Mar 18, 2026

Applicants

Alipay (Hangzhou) Information Technology Co., Ltd.

Inventors

LIU, Jian, GU, Shuwei, ZHAN, Xiaojun, RUAN, Ruoyi

Abstract

One or more embodiments of this specification provide a system for training an artificial intelligence model and a checkpoint file storage method. The system includes a model training module and a first cache module. The model training module is configured to: read a dataset needed for training from the first cache module, to execute a training task of an artificial intelligence model, where computation of the training task is performed by a GPU chip, and in a process of executing the training task, generate a checkpoint file and send the checkpoint file to the first cache module. The first cache module is configured to: identify a type of obtained to-be-stored data; and if the type of the to-be-stored data is a dataset, first write the to-be-stored data into a local buffer, and then store the to-be-stored data in a local hard disk from the local buffer; or if the type of the to-be-stored data is a checkpoint file, directly store the to-be-stored data in the local hard disk.

IPC Classifications

G06F 12/0868 20160101AFI20250508BHEP G06F 12/0888 20160101ALI20250508BHEP G06N 20/00 20190101ALI20250508BHEP

Designated States

AL, AT, BE, BG, CH, CY, CZ, DE, DK, EE, ES, FI, FR, GB, GR, HR, HU, IE, IS, IT, LI, LT, LU, LV, MC, ME, MK, MT, NL, NO, PL, PT, RO, RS, SE, SI, SK, SM, TR

View original document →