Update app.py

app.py

@@ -1,6 +1,8 @@
 # import part
 import streamlit as st
 from transformers import pipeline
+from gtts import gTTS  # Using gTTS for text-to-speech
+import os
 
 # function part
 # img2text
@@ -18,16 +20,17 @@ def text2story(text):
     # Add a prompt to guide the story generation (avoid magical elements)
     prompt = f"Write a fun and realistic story for kids based on this: {text}. Keep it simple and under 95 words."
     story = story_generator(prompt, max_length=95, num_return_sequences=1)[0]["generated_text"]
-    # Ensure the story is simple and fun for kids
-    simple_story = story[:95]  # Limit to 95 words
-    return simple_story
+    # Remove the prompt from the generated story
+    story = story.replace(prompt, "").strip()  # Clean up the output
+    return story[:95]  # Limit to 95 words
 
 # text2audio
 def text2audio(story_text):
-    # Use a TTS model to convert text to audio
-    tts = pipeline("text-to-speech", model="facebook/fastspeech2-en-ljspeech")
-    audio_data = tts(story_text)
-    return audio_data
+    # Use gTTS for text-to-speech conversion
+    tts = gTTS(text=story_text, lang="en")
+    audio_file = "story_audio.mp3"
+    tts.save(audio_file)
+    return audio_file
 
 # main part
 st.set_page_config(page_title="Story Explorer", page_icon="🦜")
@@ -54,8 +57,11 @@ if uploaded_file is not None:
 
     # Stage 3: Story to Audio data
     st.text('Turning your story into audio... 🎧')
-    audio_data = text2audio(story)
+    audio_file = text2audio(story)
 
     # Play button
     if st.button("Play Audio"):
-        st.audio(audio_data['audio'], format="audio/wav", start_time=0, sample_rate=audio_data['sampling_rate'])
+        st.audio(audio_file, format="audio/mp3")
+
+    # Clean up the audio file after use
+    os.remove(audio_file)