app.py

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

# —––––––– Requirements —–––––––
# streamlit>=1.20
# pillow>=9.0
# torch>=2.0.0
# transformers>=4.30
# sentencepiece>=0.1.97
# gTTS>=2.3.1

# —––––––– Page Setup —–––––––
st.set_page_config(page_title="Magic Story Generator", layout="centered")
st.title("📖✨ Turn Images into Children's Stories")

# —––––––– Load Pipelines (cached) —–––––––
@st.cache_resource(show_spinner=False)
def load_pipelines():
    # 1) Image-captioning pipeline (BLIP)
    captioner = pipeline(
        task="image-to-text",
        model="Salesforce/blip-image-captioning-base",
        device=-1  # CPU; set to 0+ for GPU
    )
    # 2) Story-generation pipeline (DeepSeek-R1-Distill-Qwen)
    storyteller = pipeline(
        task="text2text-generation",
        model="deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B",
        tokenizer="deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B",
        trust_remote_code=True,
        device=-1,                # CPU; set to 0+ for GPU
        temperature=0.6,
        top_p=0.9,
        repetition_penalty=1.1,
        no_repeat_ngram_size=2,
        max_new_tokens=120
    )
    return captioner, storyteller

captioner, storyteller = load_pipelines()

# —––––––– Main App —–––––––
uploaded = st.file_uploader("Upload an image:", type=["jpg", "png", "jpeg"])
if uploaded:
    # Load and display the image
    img = Image.open(uploaded).convert("RGB")
    st.image(img, use_container_width=True)

    # Generate caption
    with st.spinner("🔍 Generating caption..."):
        cap = captioner(img)
        caption = cap[0].get("generated_text", "").strip() if isinstance(cap, list) else ""
    if not caption:
        st.error("😢 Couldn't understand this image. Try another one!")
        st.stop()
    st.success(f"**Caption:** {caption}")

    # Build prompt and generate story
    prompt = (
        f"Image description: {caption}\n"
        "Write a coherent, 50-100 word children’s story that flows naturally."
    )
    with st.spinner("📝 Writing story..."):
        start = time.time()
        out = storyteller(prompt)
        gen_time = time.time() - start
        st.text(f"⏱ Generated in {gen_time:.1f}s")
    story = out[0].get("generated_text", "").strip()

    # Enforce ≤100 words
    words = story.split()
    if len(words) > 100:
        story = " ".join(words[:100]) + ("" if story.endswith('.') else ".")

    # Display story
    st.subheader("📚 Your Magical Story")
    st.write(story)

    # Convert to audio
    with st.spinner("🔊 Converting to audio..."):
        try:
            tts = gTTS(text=story, lang="en", slow=False)
            tmp = tempfile.NamedTemporaryFile(delete=False, suffix=".mp3")
            tts.save(tmp.name)
            st.audio(tmp.name, format="audio/mp3")
        except Exception as e:
            st.warning(f"⚠️ TTS failed: {e}")

# Footer
st.markdown("---\n*Made with ❤️ by your friendly story wizard*")