app.py

import streamlit as st
from transformers import pipeline

# Load model with error handling
@st.cache_resource
def load_model():
    try:
        return pipeline('sentiment-analysis', model="distilbert-base-uncased-finetuned-sst-2-english")
    except Exception as e:
        st.error(f"Model loading failed: {str(e)}")
        st.stop()

classifier = load_model()

# Streamlit UI
st.title("Sentiment Analysis")
user_input = st.text_area("Enter text to analyze:", "I love this simple version!")

if st.button("Analyze"):
    if user_input:
        try:
            result = classifier(user_input)
            st.subheader("Result:")
            emoji = "😊" if result[0]['label'] == 'POSITIVE' else "😞"
            st.write(f"{emoji} **{result[0]['label']}** (confidence: {result[0]['score']:.2%})")
        except Exception as e:
            st.error(f"Analysis failed: {str(e)}")
    else:
        st.warning("Please enter some text first!")