The reason that there's no audio available is probably that speech synthesis requires extensive data collected from the speakers of the language. There are not a lot of High Valyrian speakers to collect data from, and trying to use a rule-based model will probably just make the pronunciations artificial and inaccurate. So unless someone actually takes time to record every sentence, don't expect audio for this course any time soon.
Seeems the made up languages like this and Klingon don't have audio. Probably because the creators would disagree
It's probably because they don't have any speakers to base their TTS on. The volunteer staff could go through and manually record the audio for each sentence, but it seems nobody wants to.
I doubt George R. R. Martin or David J. Peterson would care if some course on the web for High Valyrian had (gasp) people actually speaking the language.