Update app.py

app.py

@@ -2,7 +2,7 @@ import streamlit as st
 import pdfplumber
 import re
 from transformers import AutoTokenizer, AutoModelForCausalLM, pipeline
-from tortoise.api import TTS
+from gtts import gTTS
 from sklearn.feature_extraction.text import CountVectorizer
 from nltk.sentiment import SentimentIntensityAnalyzer
 
@@ -10,7 +10,6 @@ from nltk.sentiment import SentimentIntensityAnalyzer
 tokenizer = AutoTokenizer.from_pretrained("Qwen/Qwen-7B", trust_remote_code=True)
 model = AutoModelForCausalLM.from_pretrained("Qwen/Qwen-7B", trust_remote_code=True)
 qa_pipeline = pipeline("question-answering")
-tts = TTS()
 sia = SentimentIntensityAnalyzer()
 
 # Helper functions
@@ -61,8 +60,12 @@ def answer_question_with_context(question, context, chunk_size=500):
             continue
     return " ".join(answers)
 
-def text_to_speech(text, voice="emotional_voice"):
-    return tts.generate(text, preset="standard", voice=voice)
+# Replace Tortoise-TTS with gTTS for text-to-speech functionality
+def text_to_speech(text, language="en"):
+    tts = gTTS(text=text, lang=language, slow=False)
+    file_name = "output_audio.mp3"
+    tts.save(file_name)
+    return file_name
 
 def extract_keywords(text, top_n=10):
     vectorizer = CountVectorizer(stop_words="english")
@@ -95,8 +98,8 @@ if uploaded_file:
 
         if st.button("Convert Summary to Audiobook"):
             with st.spinner("Generating audio..."):
-                audio = text_to_speech(summary)
-            st.audio(audio, format="audio/wav", start_time=0)
+                audio_file = text_to_speech(summary)
+            st.audio(audio_file, format="audio/mp3", start_time=0)
 
     st.markdown("### Ask Questions About the Document")
     question = st.text_input("Your Question:")
@@ -106,8 +109,8 @@ if uploaded_file:
         st.write(f"**Answer:** {answer}")
         if st.button("Convert Answer to Audio"):
             with st.spinner("Generating answer audio..."):
-                answer_audio = text_to_speech(answer)
-            st.audio(answer_audio, format="audio/wav", start_time=0)
+                answer_audio_file = text_to_speech(answer)
+            st.audio(answer_audio_file, format="audio/mp3", start_time=0)
 
     st.markdown("### Document Insights")
     if st.checkbox("Extract Keywords"):