Allow the configuration of a Vosk server to do speech recognition.

[Breizhux]

Hello, here's an idea that came up while discussing your application. It would be to make it possible to configure a Vosk server. Instead of doing the transcription entirely locally, it would be done remotely on a Vosk server, which would make it possible to use Vosk's much heavier templates, which are also much better. And of course, keep the default usage locally (not all users will know how to configure their own Vosk server, I'm well aware of that).

Otherwise, thank you very much for this excellent application! Good luck!