unable to load a txt file in synthesizer preprocessing.

[hanaa-salah]

I'm having a problem with the function of the preprocess_speaker in synthesizer, especially in the part of getting the corresponding text
I use Data of Librispeech, but inside the data, each speaker has only one text file
In the code, it needs a file that ends with .txt in data. The file of all audios of one speaker ends with .trans.text and also no file named .normalized.txt in data
is there is any solution for this ?