app.py

import streamlit as st
from PIL import Image
from transformers import pipeline
from gtts import gTTS
import tempfile
import os

# —––––––– Page config —–––––––
st.set_page_config(
    page_title="Storyteller for Kids",
    page_icon="📚",
    layout="centered",
    initial_sidebar_state="collapsed"
)
st.title("🖼️➡️📖 Interactive Storyteller")

# —––––––– Cache model loading —–––––––
@st.cache_resource
def load_pipelines():
    # Image-to-text pipeline
    captioner = pipeline(
        "image-to-text",
        model="Salesforce/blip-image-captioning-base",
        max_new_tokens=50
    )
    
    # Story generation pipeline with better parameters
    storyteller = pipeline(
        "text2text-generation",
        model="google/flan-t5-xxl",
        device_map="auto",
        model_kwargs={"load_in_8bit": True}
    )
    
    return captioner, storyteller

# —––––––– Main workflow —–––––––
def main():
    captioner, storyteller = load_pipelines()
    
    # —––––––– Image upload —–––––––
    uploaded = st.file_uploader(
        "Upload an image:",
        type=["jpg", "jpeg", "png"],
        help="Max size: 5MB"
    )
    
    if uploaded:
        try:
            # —––––––– Display image —–––––––
            image = Image.open(uploaded).convert("RGB")
            st.image(image, caption="Your Image", use_column_width=True)

            # —––––––– Generate caption —–––––––
            with st.spinner("🔍 Analyzing image content..."):
                cap_outputs = captioner(image)
                cap = cap_outputs[0].get("generated_text", "").strip()
                
            st.subheader("Image Understanding")
            st.info(f"**Detected:** {cap}")

            # —––––––– Generate story —–––––––
            st.subheader("Story Creation")
            prompt = f"""Create a children's story (3-10 years old) based on this description:
            
            {cap}
            
            Requirements:
            - 50-100 words
            - Playful and imaginative
            - Positive message
            - Simple vocabulary
            - Include animal characters
            
            Story:"""
            
            with st.spinner("✍️ Crafting a magical story..."):
                story_output = storyteller(
                    prompt,
                    max_length=300,
                    do_sample=True,
                    top_p=0.95,
                    temperature=0.85,
                    num_beams=4,
                    repetition_penalty=1.2
                )
                story = story_output[0]["generated_text"].strip()
                
            st.success("**Generated Story:**")
            st.write(story)

            # —––––––– Text-to-Speech —–––––––
            st.subheader("Audio Version")
            with st.spinner("🔊 Generating audio..."):
                try:
                    tts = gTTS(text=story, lang="en", slow=False)
                    with tempfile.NamedTemporaryFile(suffix=".mp3", delete=False) as tmp:
                        tts.write_to_fp(tmp)
                        tmp_path = tmp.name
                        
                    st.audio(tmp_path, format="audio/mp3")
                    
                    # Add download button
                    with open(tmp_path, "rb") as f:
                        st.download_button(
                            label="Download Audio Story",
                            data=f,
                            file_name="kids_story.mp3",
                            mime="audio/mpeg"
                        )
                    
                finally:
                    if os.path.exists(tmp_path):
                        os.remove(tmp_path)

        except Exception as e:
            st.error(f"Error processing your request: {str(e)}")

if __name__ == "__main__":
    main()