Regarding wav files creation

I have doubt

Suppose, in wav file contain silence is good?

Ex: my sentence is how are you and wave file is 6sec so, after sentence, there is silence in end as well as start

is it required to remove that silence ?

For training or for inference?

You could try to cut out silence with tools like this

It is not required. For training, it may be beneficial to trim silence specially if you have a lot of it, because it’ll use more GPU memory.

@baconator For training the model

in inference silence or pause will be their