Model training method and device, computer equipment and storage medium

Abstract

The invention provides a model training method and device, computer equipment and a storage medium. The method comprises the steps of obtaining a sample text for model training; calling a sample multitask translation model, wherein the sample multitask translation model comprises a sample encoder, a first sample decoder and a second sample decoder; encoding the sample text based on a sample encoder to obtain text features of the sample text; decoding the text features based on a first sample decoder to obtain a first prediction text of the sample text, and decoding the text features based on asecond sample decoder to obtain a second prediction text of the sample text; and training a sample multi-task translation model according to the sample text, the first prediction text and the secondprediction text to obtain a multi-task translation model. By training the multi-task translation model, the accuracy of model translation can be improved.

Description

Technical field[0001]This application relates to the field of Internet technology, in particular to a model training method, device, computer equipment and storage medium.Background technique[0002]With the continuous development and evolution of deep learning, neural network models have been widely used in fields such as natural language processing, speech recognition, and even computer vision, such as neural network machine translation, natural language understanding, automatic speech recognition, target detection and other practical applications It is widely used.[0003]Neural network models are used in neural network machine translation, and mainly include autoregressive translation models and non-autoregressive translation models. The autoregressive translation models may specifically include Transformer models, and the non-autoregressive translation models may specifically include Mask-Predict models. In the autoregressive translation model, the translation is generated word by ...

AI Technical Summary

Problems solved by technology

[0004] In the prior art, the autoregressive translation model (or non-autoregressive translation model) includes an encoder and a decoder, and when training the regression translation model (or non-autoregressive translation model), it is based on an encoder and a decoder The model structure is single, and the translation ability learned by the model is limited, so the translation accuracy of the trained model is not high

Images