Method for obtaining prediction model, voice waveform prediction method and related device

The invention discloses a method for obtaining a prediction model, a voice waveform prediction method, electronic equipment and a computer readable storage medium, and the method for obtaining the prediction model comprises the steps: firstly dividing a sample voice waveform into a plurality of sample subsequences, carrying out the time delay processing, and then constructing an initial prediction model, and training an initial prediction model according to the acoustic parameters of the plurality of sample subsequences and the sample voice waveforms to obtain a prediction model. Wherein the waveform values of the current waveform points in the plurality of sample subsequences are input into the prediction model at the same time, and the predicted waveform values of the next waveform points in the plurality of sample subsequences can be obtained at the same time. Therefore, the calculation amount for predicting and generating the voice waveform can be reduced, and the efficiency of generating the voice wavefo