Update app.py

app.py

@@ -93,7 +93,8 @@ def generate_base(subject, setting):
     speech_output = [output.cpu().numpy() for output in speech_output]
     gr.Info("Generated Audio")
     return None, None, {"audio": speech_output, "text": model_input_tokens}
-    
+
+import time
 def stream_audio(state):
     speech_output = state["audio"]
     sentences = state["text"]
@@ -105,6 +106,7 @@ def stream_audio(state):
         print(f"Sample of length: {round(new_audio.shape[0] / sampling_rate, 2)} seconds")
         story += f"{sentence}\n"
         yield story, numpy_to_mp3(new_audio, sampling_rate=sampling_rate)
+        time.sleep(1)
 
 
 with gr.Blocks() as block: