Update modules/vad/silero_vad.py

modules/vad/silero_vad.py

@@ -256,9 +256,9 @@ class SileroVAD:
         ts_map = SpeechTimestampsMap(speech_chunks, sampling_rate)
 
         for segment in segments:
-            if segment["words"]:
+            if segment["text"]:
                 words = []
-                for word in segment["words"]:
+                for word in segment["text"]:
                     # Ensure the word start and end times are resolved to the same chunk.
                     middle = (word.start + word.end) / 2
                     chunk_index = ts_map.get_chunk_index(middle)
@@ -268,7 +268,7 @@ class SileroVAD:
 
                 segment["start"] = words[0].start
                 segment["end"] = words[-1].end
-                segment["words"] = words
+                segment["text"] = words
 
             else:
                 segment["start"] = ts_map.get_original_time(segment["start"])