[Feature request] is it possible to initialize speaker_wav 1 time?

[Vubni]

I haven't found if there is such a function, so I suggest adding it so that the cloned voice can initialize only 1 time.
I think this will significantly reduce the response generation time, because he will not have to view the audio of the cloned voice again, he will already know it

[stale]

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions. You might also look our discussion channels.