No tts on snips! how do I turn it on?

#1

Hello there, i followed the snips official tutorial for raspberry pi3 installation but I cant get to work tts functionalities!
What do I have to do? I just would like to have some easy speak like “message received, I’ll do!” or “I’m turning on the lights!” or something like that…

ps. I would like to stay offline, so no google cloud or something like that!
thank you everybody…

#2

Hi,

At minimum you need to provide the output of snips-watch -vvv taken during an interaction with your assistant to allow others to help you debug your problem.

#3

Hello
I recommend you read this thread:


and have a good read of the dialogue api reference here:
https://docs.snips.ai/reference/dialogue
But the baseline is you need to publish the things you want the snips tts to say to hermes over mqtt as part of a json object under text to either “hermes/dialogueManager/startSession”, “/continueSession”, “/endSession”.
The Json also needs to include a sessionId. So if you want to end a session after a user interaction you could do it for example by publishing this:

{"sessionId":"1234yoursessionID","text":"the text snips should say"}

to “hermes/dialogueManager/endSession”