Update app.py

app.py

@@ -2,6 +2,9 @@ import streamlit as st
 from gtts import gTTS 
 from io import BytesIO
 from PyPDF2 import PdfReader
+from whisperspeech.pipeline import Pipeline
+
+pipe = Pipeline(torch_compile=True)
 
 st.image('OIG3 (4).jpeg', caption='Your host on this PDF-to-Speech adventure!')
 
@@ -22,10 +25,19 @@ if uploaded_file is not None:
         # extracting text from page
         text = page.extract_text()  
         print("Created text of page", i )
-        sound_file = BytesIO()
-        tts = gTTS(text, lang='en')
-        tts.write_to_fp(sound_file)
-        st.audio(sound_file)
+
+    # Generate audio for the current page using a unique filename
+        page_audio_file = f"output_{i}.wav"
+        pipe.generate_to_file(page_audio_file, text)
+
+    # Display the generated audio using st.audio
+        with open(page_audio_file, "rb") as audio_file:
+            st.audio(audio_file)
+        
+#        sound_file = BytesIO()
+#        tts = gTTS(text, lang='en')
+#        tts.write_to_fp(sound_file)
+#        st.audio(sound_file)
         print("Read aloud", i, "pages of", X, "total pages.")
         i = i + 1
     st.write("🎉 That's the whole PDF! Have an awesome day! 🎉")