Training machine learning models for multiple machine learning tasks

摘要

A method of training a machine learning model having a plurality of parameters, wherein the machine learning model is trained on a first machine learning task to determine a first parameter value of the machine learning model. Is done. The method includes determining, for each of the plurality of parameters, a respective measure of the importance of the parameter to the machine learning model that achieves acceptable performance with respect to the first machine learning task, and a second different machine learning task. Obtaining training data for training the machine learning model with respect to the second machine learning task, while maintaining an acceptable level of performance for the first machine learning task. Training the machine learning model with respect to the second machine learning task by training the machine learning model with respect to the training data to adjust the first parameter value to achieve performance.