app.py

import shap
import numpy as np
import torch
from transformers import pipeline, AutoModelForSequenceClassification, AutoTokenizer
import streamlit as st

model_name = "mavinsao/mi-roberta-base-finetuned-mental-health"
model = AutoModelForSequenceClassification.from_pretrained(model_name)
tokenizer = AutoTokenizer.from_pretrained(model_name)

device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
model.to(device)

# Create a pipeline with the model and tokenizer
classifier = pipeline("text-classification", model=model, tokenizer=tokenizer)

# Streamlit app
st.title("SHAP Explanation for Mental Illness Prediction")

# Input text area for user input
text = st.text_area("Enter a sentence to explain:")

if st.button("Explain"):
    # Generate the SHAP explainer
    explainer = shap.Explainer(classifier, masker=tokenizer)

    # Compute SHAP values
    shap_values = explainer([text])

    # Save SHAP plot as HTML
    shap_html = shap.plots.text(shap_values, display=False)

    # Save the plot to an HTML file
    with open("shap_plot.html", "w") as f:
        f.write(shap_html)
    
    # Display the SHAP plot in Streamlit using components
    st.components.v1.html(shap_html, height=500, scrolling=True)