Add dictionary of common words (especially jargon)

It would be really nice if I could create a dictionary that can be used by the speech-to-text model to give a higher probability to words that are jargon or refer to certain named entities. It seems like sometimes has some trouble transcribing these words.