Spaces:
Sleeping
Sleeping
| import streamlit as st | |
| from transformers import pipeline | |
| from wordcloud import WordCloud | |
| import matplotlib.pyplot as plt | |
| from huggingface_hub import login | |
| hf_token = st.secrets.get("HF_TOKEN") | |
| if hf_token: | |
| login(hf_token) | |
| st.set_page_config(page_title="📰 News Veracity Checker", layout="wide") | |
| def load_models(): | |
| veracity_model = pipeline('text-classification', model='jy46604790/Fake-News-Bert-Detect') | |
| summarizer = pipeline('summarization', model='facebook/bart-large-cnn') | |
| explainer = pipeline('text-generation', model='google/gemma-2b-it', max_new_tokens=200) | |
| return veracity_model, summarizer, explainer | |
| veracity_model, summarizer, explainer = load_models() | |
| def get_summary(text): | |
| summary = summarizer(text, max_length=120, min_length=30, do_sample=False) | |
| return summary[0]['summary_text'] | |
| def generate_explanation(text, veracity_label): | |
| prompt = (f"You are a news analyst. Given this article:\n\n{text}\n\n" | |
| f"Explain clearly why this news is {veracity_label.lower()}. " | |
| f"If false, provide an accurate version or correct factual information. Explanation:") | |
| response = explainer(prompt)[0]['generated_text'] | |
| return response.split("Explanation:")[-1].strip() | |
| def plot_wordcloud(text): | |
| wc = WordCloud(width=800, height=400, background_color='white').generate(text) | |
| plt.figure(figsize=(10, 5)) | |
| plt.imshow(wc, interpolation='bilinear') | |
| plt.axis('off') | |
| st.pyplot(plt) | |
| def get_veracity(text): | |
| res = veracity_model(text[:512])[0] | |
| label = res['label'] | |
| score = round(res['score'], 3) | |
| if label == "LABEL_0": | |
| return "False", score | |
| else: | |
| return "True", score | |
| #Streamlit | |
| st.title("📰 News Veracity Checker") | |
| st.sidebar.info("Paste a news headline or article to get a veracity check, summary, and AI-generated explanation.") | |
| input_text = st.text_area("Paste news headline or article:", height=200) | |
| if st.button("🚀 Analyze"): | |
| with st.spinner("Analyzing the content..."): | |
| label, score = get_veracity(input_text) | |
| if label == "True": | |
| st.markdown(f"""<h4>✅ <span style='color:green;'>Veracity: True</span> ({score*100:.2f}% confidence)</h4>""",unsafe_allow_html=True) | |
| else: | |
| st.markdown(f"""<h4>❌ <span style='color:red;'>Veracity: False</span> ({score*100:.2f}% confidence)</h4>""",unsafe_allow_html=True) | |
| #summary | |
| summary = get_summary(input_text) | |
| st.markdown("### 📃 **Summary:**") | |
| st.info(summary) | |
| #explanation | |
| explanation = generate_explanation(input_text, label) | |
| st.markdown("### 🤖 **AI Explanation:**") | |
| st.write(explanation) | |
| plot_wordcloud(input_text) | |