apptest.py

"""
Magic Story Generator App for Hugging Face Spaces
Creates custom children's stories from uploaded images
"""

import streamlit as st
from PIL import Image
import tempfile
import numpy as np
from transformers import pipeline
import torch
import os

# ======================
# UI Configuration
# ======================
def configure_ui():
    """Sets up child-friendly interface with custom styling"""
    st.set_page_config(
        page_title="✨ Magic Story Generator",
        page_icon="🧚",
        layout="wide"
    )
    
    # Custom CSS for child-friendly design
    st.markdown("""
    <style>
    .main {
        background-color: #FFF5E6;
        background-image: url('https://img.freepik.com/free-vector/hand-drawn-childish-pattern_23-2149073136.jpg');
        background-size: 30%;
        opacity: 0.9;
    }
    h1 {
        color: #FF6B6B;
        font-family: 'Comic Sans MS', cursive;
        text-align: center;
        text-shadow: 2px 2px 4px #FFD166;
    }
    .stButton>button {
        background-color: #4ECDC4;
        color: white;
        border-radius: 20px;
        padding: 10px 24px;
        font-weight: bold;
    }
    .stFileUploader>div>div>div>div {
        border: 2px dashed #FF9E7D;
        border-radius: 15px;
        background-color: #FFF0F5;
    }
    .story-box {
        background-color: #FFF0F5;
        padding: 20px;
        border-radius: 15px;
        border-left: 5px solid #FF6B6B;
        font-family: 'Comic Sans MS', cursive;
        font-size: 18px;
        line-height: 1.6;
    }
    </style>
    """, unsafe_allow_html=True)

# ======================
# Stage 1: Image Captioning
# ======================
@st.cache_resource
def load_image_captioner():
    """Loads BLIP image captioning model with GPU support if available"""
    return pipeline(
        "image-to-text",
        model="Salesforce/blip-image-captioning-base",
        device="cuda" if torch.cuda.is_available() else "cpu"
    )

def generate_caption(_pipeline, image):
    """Generates English description of uploaded image"""
    try:
        # Generate caption with 20-50 words
        result = _pipeline(image, max_new_tokens=50)
        return result[0]['generated_text']
    except Exception as e:
        st.error(f"Caption generation failed: {str(e)}")
        return None

# ======================
# Stage 2: Story Generation
# ======================
@st.cache_resource
def load_story_generator():
    """Loads fine-tuned GPT-2 story generator"""
    return pipeline(
        "text-generation",
        model="pranavpsv/gpt2-genre-story-generator",
        device="cuda" if torch.cuda.is_available() else "cpu"
    )

def generate_story(_pipeline, keywords):
    """Creates a children's story (60-100 words) based on image caption"""
    prompt = f"""Generate a children's story (60-100 words) based on: {keywords}
    Requirements:
    - Use simple English (Grade 2 level)
    - Include magical/fantasy elements
    - Have positive moral lesson
    - Happy ending
    - Exactly 3 paragraphs
    
    Story:"""
    
    try:
        story = _pipeline(
            prompt,
            max_length=250,  # Controls token count (~100 words)
            temperature=0.7, # Balance creativity vs coherence
            do_sample=True,
            top_k=50
        )[0]['generated_text']
        # Clean up output by removing prompt
        return story.replace(prompt, "").strip()
    except Exception as e:
        st.error(f"Story generation failed: {str(e)}")
        return None

# ======================
# Stage 3: Text-to-Speech
# ======================
@st.cache_resource
def load_tts():
    """Loads multilingual TTS model"""
    return pipeline(
        "text-to-speech",
        model="facebook/mms-tts-eng",
        device="cuda" if torch.cuda.is_available() else "cpu"
    )

def text_to_speech(_pipeline, text):
    """Converts generated story to speech audio"""
    try:
        audio = _pipeline(text)
        with tempfile.NamedTemporaryFile(delete=False, suffix=".wav") as f:
            import soundfile as sf
            sf.write(f.name, audio["audio"].squeeze().numpy(), audio["sampling_rate"])
            return f.name
    except Exception as e:
        st.error(f"Audio generation failed: {str(e)}")
        return None

# ======================
# Main Application
# ======================
def main():
    # Configure UI first
    configure_ui()
    
    # App header
    st.title("🧚 Magic Story Generator")
    st.markdown("""
    <div style="text-align:center; color:#FF8E72; font-family: 'Comic Sans MS'; font-size: 20px;">
    Upload a child's photo and AI will create a custom fairy tale with audio!
    </div>
    """, unsafe_allow_html=True)
    
    # File upload section
    uploaded_file = st.file_uploader(
        "Choose a photo of children's activity",
        type=["jpg", "jpeg", "png"],
        help="Examples: playing, reading, drawing etc."
    )
    
    if not uploaded_file:
        st.info("👆 Please upload an image to begin")
        return
    
    # Display uploaded image
    image = Image.open(uploaded_file)
    st.image(image, caption="Your uploaded photo", use_column_width=True)
    
    # Load all models (shows loading animation)
    with st.spinner("🪄 Preparing magic tools..."):
        caption_pipe = load_image_captioner()
        story_pipe = load_story_generator()
        tts_pipe = load_tts()
    
    # --- Stage 1: Image Captioning ---
    with st.spinner("🔍 Analyzing the image..."):
        caption = generate_caption(caption_pipe, image)
        if caption:
            st.success(f"📝 AI sees: {caption}")
    
    # --- Stage 2: Story Generation ---
    if caption:
        with st.spinner("✍️ Writing your story..."):
            story = generate_story(story_pipe, caption)
            if story:
                st.subheader("📖 Your Custom Story")
                st.markdown(f"""
                <div class="story-box">
                {story}
                </div>
                """, unsafe_allow_html=True)
                
                # --- Stage 3: Text-to-Speech ---
                with st.spinner("🔊 Creating audio version..."):
                    audio_path = text_to_speech(tts_pipe, story)
                    if audio_path:
                        st.audio(audio_path, format="audio/wav")
                        st.success("Audio ready! Click play above to listen")
                        st.balloons()  # Celebration animation

if __name__ == "__main__":
    # Set Hugging Face cache location
    os.environ["HF_HUB_CACHE"] = "/tmp/huggingface"
    main()