-
Notifications
You must be signed in to change notification settings - Fork 30
Open
Labels
kind/questionFurther information is requestedFurther information is requestedtriage/needs-investigationIssue needs investigation before it can be confirmed.Issue needs investigation before it can be confirmed.
Description
I've connected together ChatGPT, Eleven Labs, and Assembly AI via everyone's streaming APIs. I'm running into the situation where the TTS is getting piped back into my microphone, which gets transcribed by Assembly AI, and fed back into ChatGPT. Essentially my AI bot gets caught in a loop, infinitely talking to itself. It's actually pretty funny, but not all that useful.
I haven't been able to find a straightforward to mute the mic directly via pyaudio
. Can we have a way to toggle transcription input?
Metadata
Metadata
Assignees
Labels
kind/questionFurther information is requestedFurther information is requestedtriage/needs-investigationIssue needs investigation before it can be confirmed.Issue needs investigation before it can be confirmed.