Deep learning model training method

Abstract

The invention discloses a deep learning model training method. The method comprises the steps of obtaining a first data set; setting an initial model parameter and a training suspension condition of the deep learning model, performing multiple times of iterative training on the deep learning model based on the first data set, and calculating a model parameter value generated by each time of iteration according to the initial model parameter; when the training suspension condition is met, generating a model node; and storing the model parameter value generated by the last iteration training as the node information of the model node. According to the deep learning model training method provided by the invention, at least two model nodes can be formed in the training process, and node information is stored. Furthermore, a display interface of the node information can be formed, and interactive operation of deep learning model training is provided on the display interface, so that a user can perform timely adjustment according to the node information of each model node in the training process to obtain an optimal deep learning model more quickly.

Description

technical field [0001] The present invention relates to the technical field of deep learning, in particular to a deep learning model training method. Background technique [0002] With the increasing maturity of AI (Artificial Intelligence) technology and theory, the application fields of AI algorithms are becoming wider and wider, and the business volume is showing a trend of rapid development. In order to free algorithm engineers from tedious engineering development tasks and concentrate on algorithm research and development, various machine learning algorithm platforms have emerged. In the existing technology, most AI algorithm platforms only provide a list of training records sorted by time, the training process cannot be managed, and the interaction with users during the model training process is poor. [0003] Therefore, the prior art still needs to be improved and developed. Contents of the invention [0004] The technical problem to be solved by the present inven...

AI Technical Summary

Problems solved by technology

[0004] The technical problem to be solved by the present invention is to provide a deep learning model training method aimed at solving the problem of poor interaction with users during the model training process in the prior art.

Images