Streamed Text Input to Whisper TTS

[saduri2004]

I know that there is a way to stream the audio file output, as seen in the Whisper Docs, but I was wondering if anyone has found a way (or if it is even possible to engineer a way) to stream the input. I am taking the output of the completions API call (where streamed output is set to TRUE), so as the stream comes in from completions endpoint, I want to speak it out at the same time. Is this feasible?

[makaveli10]

@saduri2004 give WhisperLive a try it supports streaming to the server running Whisper model.

[jayavanth]

@makaveli10 curious how it works. Is it chunking the input?