app.py

# import part
import streamlit as st
from transformers import pipeline
import textwrap
import numpy as np
import soundfile as sf
import tempfile
import os
from PIL import Image
import string

# Initialize pipelines with caching
@st.cache_resource
def load_pipelines():
    captioner = pipeline("image-to-text", model="Salesforce/blip-image-captioning-large")
    storyer = pipeline("text-generation", model="gpt2")
    tts = pipeline("text-to-speech", model="facebook/mms-tts-eng")
    return captioner, storyer, tts

captioner, storyer, tts = load_pipelines()

# Function part
# Function to generate content from an image
def generate_content(image):
    pil_image = Image.open(image)
    
    # Generate caption
    caption = captioner(pil_image)[0]["generated_text"]
    st.write("**🌟 What's in the picture: 🌟**")
    st.write(caption)

# Create prompt for story
prompt = (
    f"Write a funny, interesting children's story centered on this scene: {caption}\n"
    f"Story in third-person narrative, describing this scene exactly: {caption} "
    f"Mention the exact place, location, or venue within {caption}. "
    f"Avoid numbers, random letter combinations, and single-letter words.")

# Generate raw story with optimized parameters
raw = storyer(
    prompt,
    max_new_tokens=100,  
    temperature=0.6,    
    top_p=0.85,          
    no_repeat_ngram_size=0, 
    return_full_text=False
)[0]["generated_text"].strip()

# Combine cleaning and word trimming in one step
# Use regex to keep only allowed characters and remove single-letter words
allowed_pattern = re.compile(r'[a-zA-Z0-9.,!?"\'-]+\b(?<!\b\w\b)')
clean_raw = ' '.join(word for word in re.findall(allowed_pattern, raw) if len(word) > 1)

def generate_story(raw, caption, tts):  
    # Split into words and trim to 100 words
    words = raw.split()  
    story = " ".join(words[:100])
    
    # Clean the story using clean_generated_story
    story = clean_generated_story(raw)  
    
    # Display story in Streamlit
    st.write("**📖 Your funny story: 📖**")
    st.write(story)
    
    # Generate audio from cleaned story
    chunks = textwrap.wrap(story, width=200)
    audio = np.concatenate([tts(chunk)["audio"].squeeze() for chunk in chunks])

    # Save audio to temporary file
    with tempfile.NamedTemporaryFile(delete=False, suffix=".wav") as temp_file:
        sf.write(temp_file.name, audio, tts.model.config.sampling_rate)
        temp_file_path = temp_file.name

    return caption, story, temp_file_path

# Streamlit UI
st.title("😎Story Maker")
st.markdown("Upload a picture, I will generate a story for you")

uploaded_image = st.file_uploader("Choose your picture", type=["jpg", "jpeg", "png"])

# Streamlit UI (modified image display section)
if uploaded_image is None:
    st.image("https://example.com/placeholder_image.jpg", caption="Upload your picture here!", use_container_width=True)
else:
    st.image(uploaded_image, caption="Your Picture ", use_container_width=True)

if st.button("Generate a story"):
    if uploaded_image is not None:
        with st.spinner("Processing"):
            caption, story, audio_path = generate_content(uploaded_image)
            st.success(" Your story is ready!😊")
            st.audio(audio_path, format="audio/wav")
            os.remove(audio_path)
    else:
        st.warning("Please upload a picture first! ")