app.py

import json
import urllib

import streamlit as st
from transformers import pipeline


TEST_METRICS = 'https://huggingface.co/nbroad/longformer-base-health-fact/raw/main/test_results.json'
st.title('Veracity of a claim with pretrained longformer (PUBHEALTH)')
st.header('Claim')
claim = st.text_area(label='Enter a claim text in the box')
pl = pipeline('text-classification',
              model='nbroad/longformer-base-health-fact')
st.header('Results')
st.write(pl(claim))
st.header('Self-reported test metrics')
with urllib.request.urlopen(TEST_METRICS) as response:
    metrics = json.load(response)
st.write(metrics)