Deep learning model training method and related equipment

• Main Authors: SU GUIYUAN, YANG ZUHUA, LIU SIYANG
• Format: Patent
• Language: Chinese; English
• Published: 30.01.2024
• Subjects: CALCULATING; COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS; COMPUTING; COUNTING; ELECTRIC DIGITAL DATA PROCESSING; PHYSICS

The embodiment of the invention provides a deep learning model training method and related equipment, which are used for protecting data security as much as possible while meeting large model training requirements. The method provided by the embodiment of the invention comprises the steps of segmenting the deep learning model into a first deep learning sub-model and a second deep learning sub-model by taking a first irreversible operation layer in the deep learning model as a segmentation point; deploying the first deep learning sub-model in a trusted execution environment, and deploying the second deep learning sub-model in a non-trusted execution environment; acquiring initial training data of a data provider, and inputting the initial training data into the first deep learning sub-model to obtain intermediate output data; transmitting the intermediate output data to a second deep learning sub-model to obtain training result data corresponding to the second deep learning sub-model; and according to the trai