this.deepgramLive = deepgram.transcription.live({
encoding: "mulaw",
sample_rate: "8000",
model: "nova-2",
punctuate: false,
interim_results: false,
smart_format: false,
endpointing: 100,
utterance_end_ms: 1000,
});
can you suggest a better configuration so that i can receive the transcription in the most quickest possible way, currently it sends 4-5 atlernate response before sending is_final
@fluid yoke @little kiln