Update app.py

app.py

@@ -294,7 +294,7 @@ def transcribe(stream, new_chunk):
                 print(audio_chunk.shape)
                 print("--- %s seconds ---" % (time.time() - start_time))
                 # yield (stream,output, audio_output)
-                time.sleep((len(audio_chunk)/text2speech.fs)-0.2)
+                # time.sleep((len(audio_chunk)/text2speech.fs)-0.2)
                 curr_output = ""
             generated_text = output
             
@@ -314,7 +314,7 @@ def transcribe(stream, new_chunk):
     #     audio_output=None
     text_str1=text_str
     
-    return (stream,text_str1, audio_output)
+    yield (stream,text_str1, audio_output)
 
 demo = gr.Interface(
     transcribe,