Today the STT returns only “did not understand” over the wyoming response. Hence I as a user do not know what went wrong…
It would be great if the response would include what is also shown in the INFO on the server. Meaning what the STT model has returned.
This way we could tell the user: hey, we understood the following…maybe you have to pronounce it a little better
And in a far future, we could use this to ask the user, if he would like to train the model with his input…
any idea where to place this feature request?