app.py

import streamlit as st
from transformers import pipeline

# Load the sentiment analysis model from Hugging Face
@st.cache_resource  # Cache the model for faster loading
def load_model():
    return pipeline('sentiment-analysis', model='cardiffnlp/twitter-roberta-base-sentiment')

model = load_model()

# Streamlit UI
st.title("Sentiment Analysis App using GenAI Models")

# Text input from the user
user_input = st.text_area("Enter text to analyze sentiment:")

# Prediction button
if st.button("Analyze"):
    if user_input:
        # Perform prediction
        result = model(user_input)  # Hugging Face pipeline returns a dictionary
        sentiment = result[0]['label']  # Get sentiment label (Positive/Negative/Neutral)
        confidence = result[0]['score']  # Confidence score
        
        # Display the sentiment and confidence score
        st.write(f"**Predicted Sentiment:** {sentiment}")
        st.write(f"**Confidence Score:** {confidence:.2f}")
    else:
        st.warning("Please enter some text to analyze.")

# Optional: Footer
st.write("---")
st.caption("Built with Streamlit and Hugging Face's GenAI models.")