File size: 184 Bytes
5fa1a76
 
1
2
Create a prepare_dataset function that takes in a 
single example and uses the SpeechT5Processor object to tokenize the input text and load the target audio into a log-mel spectrogram.