Update app.py

app.py

@@ -9,7 +9,7 @@ def get_labels(audio_fp, threshold, min_speech_duration_ms, min_silence_duration
     audio = AudioSegment.from_file(audio_fp)
     speech_timestamps = detect_nonsilent(audio, min_silence_len=min_silence_duration_ms, silence_thresh=-40)
     speech_timestamps = list(filter(lambda x: x[1]-x[0] > min_speech_duration_ms, speech_timestamps))
-    speech_timestamps = [{"start": s[0], "end": s[1]} for s in speech_timestamps]
+    speech_timestamps = [{"start": s[0]/1000, "end": s[1]/1000} for s in speech_timestamps]
     labels_str = []
     labels = []