Closed ethanzrd closed 1 year ago
Upgraded the system with VAD, ensuring that only speech-containing chunks make it into the queue.
Note: with a 5-second transcription timeout, the transcription triggers for every 5 seconds of speech, not every 5 seconds.
Fixed end-of-stream behavior for the sequential mode.
Upgraded the system with VAD, ensuring that only speech-containing chunks make it into the queue.
Note: with a 5-second transcription timeout, the transcription triggers for every 5 seconds of speech, not every 5 seconds.
Fixed end-of-stream behavior for the sequential mode.