Add big file recognition demo

[vadimdddd]

Issue https://github.com/alphacep/vosk-api/issues/1117. How to run:

1. You have to download small and big model for your language and place it here Path.home() .cache/vosk - it will be done auto in the next commits.
2. run demo like: python3 test_big_file.py path_to_your_file processing of audio file fragments will be carried out in parallel in a large model - this will be introduced in future commits