Training machine learning models for multiple machine learning tasks

摘要

A method is disclosed for training a machine learning model having multiple parameters trained for a first machine learning task to determine first values of parameters of the machine learning model. The method comprising the steps of: for each of the plurality of parameters, determining a measure of a significance of a parameter for the machine learning model to achieve an acceptable performance for the first machine learning task; Obtaining training data for training the machine learning model for a second different machine learning task; And training the machine learning model for the training data so that the machine learning model achieves an acceptable level of performance for a second machine learning task while maintaining an acceptable level of performance for the first machine learning task, And training the machine learning model for a second machine learning task by adjusting the first values of the parameters.