Customizing voice / tone generation with specific input vocal

[Tortoise17]

Dear Friends,

Since, I am trying to understand the working of this tool. I want to ask audioldm2 -t "A female reporter is speaking full of emotion" --transcription "Wish you have a good day"

This is simple example which generates the audio of the female speaker which is random voice. Can we customize it to specific voice by giving input audio small file as input sequence? I have seen input sequence function. But still I could not understand how to adopt it.? Please guide.