How can I use a phonetic dictionary (only)

Hi!

I currently try to figure out how to use Vosk (for NodeJS. Thanks for the cool demos!) with a phonetic dictionary (maybe instead of a full language model?). My use case is to recognize if the speaker says one certain keyword like "Anton". In the past we used a phonetic dictionary and keyword file with PocketSphinx. I´ve read the documentation [here](https://alphacephei.com/vosk/lm) but could not find a hint in the API to achieve that.

Any help would be great,
Thank you!