Update app.py

app.py

@@ -72,12 +72,9 @@ def centered_text(input_text, mode = "h1",):
 def generate_voice(input_text):
     # TTS Inference
     voice = st.session_state.TTS.inference(input_text)
-
-    # Save audio (bug in Streamlit, can't play numpy array directly)
-    sf.write(f"cache_sound/new_generated.wav", voice, 22050)
     
     # Play audio
-    st.audio(f"cache_sound/new_generated.wav", format = "audio/wav")
+    st.audio(voice,sample_rate = 22050)
     os.remove(f"cache_sound/new_generated.wav")
     st.caption("Generated Voice")