Is your feature request related to a problem? Please describe.

Parakeet TDT 0.6B V3 reportedly handles 25 languages, which is awesome. However, when transcribing an unknown audio, I'd like to be able to know what language it is.

Describe the solution you'd like

Is it possible to include the code of the language detected by parakeet in the output ?
For instance a "language" property with values "en", "fr", "es", ...
Or is there any other way to achieve it currently ?
I was not able to find any related API documentation.

Context

I am serving Parakeet as a transcription service using Triton inference server.