Initial UI

.gitignore

@@ -0,0 +1,54 @@
+# Python
+__pycache__/
+*.py[cod]
+*$py.class
+*.so
+.Python
+build/
+develop-eggs/
+dist/
+downloads/
+eggs/
+.eggs/
+lib/
+lib64/
+parts/
+sdist/
+var/
+wheels/
+*.egg-info/
+.installed.cfg
+*.egg
+MANIFEST
+
+# Virtual environments
+.env
+.venv
+env/
+venv/
+ENV/
+env.bak/
+venv.bak/
+
+# IDE
+.vscode/
+.idea/
+*.swp
+*.swo
+*~
+
+# OS
+.DS_Store
+.DS_Store?
+._*
+.Spotlight-V100
+.Trashes
+ehthumbs.db
+Thumbs.db
+
+# Gradio
+gradio_cached_examples/
+flagged/
+
+# Hugging Face
+.huggingface/ 

README.md

@@ -1,14 +1,55 @@
 ---
-title: Hf News
-emoji: 🐠
-colorFrom: green
-colorTo: blue
+title: HF News (AI-assisted summary)
+emoji: 📰
+colorFrom: blue
+colorTo: purple
 sdk: gradio
-sdk_version: 5.38.1
+sdk_version: 4.44.0
 app_file: app.py
 pinned: false
 license: apache-2.0
-short_description: Aggregated Hugging Face News
+short_description: Stay updated with AI/ML news, summarized by AI
 ---
 
-Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
+# 🚀 HF News (AI-assisted summary) 📰
+
+A playful Gradio application that displays the latest AI and ML news summaries from the [HF News Aggregator dataset](https://huggingface.co/datasets/rajatarya/hf-news-aggregator).
+
+## ✨ Features
+
+- 📰 **Latest News First**: Always shows the most recent AI-summarized news
+- 🧭 **Easy Navigation**: Browse through older summaries with intuitive controls  
+- 📊 **Rich Metadata**: View details about each summary including sources, time range, and analysis stats
+- 🎨 **Playful UI**: Emoji-rich interface with a modern, friendly design
+- 🔄 **Live Updates**: Refresh button to get the latest data from the dataset
+
+## 🎯 How it Works
+
+The application fetches data from the `rajatarya/hf-news-aggregator` dataset, which contains AI-generated summaries of news from various sources. Each summary includes:
+
+- **Summarized Content**: AI-generated markdown summaries of the latest AI/ML news
+- **Metadata**: Information about the summary generation process
+- **Sources**: Details about the news sources analyzed
+- **Time Range**: The period covered by each summary
+
+## 🚀 Usage
+
+Simply visit the Space and:
+1. 👀 View the latest news summary
+2. ➡️ Navigate to newer summaries with the "Newer News" button
+3. ⬅️ Browse older summaries with the "Older News" button  
+4. 🔄 Refresh to get the latest data from the dataset
+
+## 🛠️ Technical Details
+
+Built with:
+- **Gradio**: For the web interface
+- **Pandas**: For data manipulation
+- **Hugging Face Hub**: For dataset access
+- **PyArrow**: For parquet file handling
+
+The app automatically loads the latest parquet file from the dataset and presents it in a user-friendly format with proper markdown rendering.
+
+---
+
+*🤖 Powered by AI • 📊 Data from [HF News Aggregator](https://huggingface.co/datasets/rajatarya/hf-news-aggregator)*

app.py

@@ -0,0 +1,197 @@
+import gradio as gr
+import pandas as pd
+from huggingface_hub import hf_hub_download
+import json
+from datetime import datetime
+import os
+
+def load_data():
+    """Load the latest news data from the parquet file"""
+    try:
+        file_path = hf_hub_download(
+            repo_id='rajatarya/hf-news-aggregator', 
+            filename='data/news_data_summaries.parquet', 
+            repo_type='dataset'
+        )
+        df = pd.read_parquet(file_path)
+        # Sort by date_generated descending to get latest first
+        df = df.sort_values('date_generated', ascending=False).reset_index(drop=True)
+        return df
+    except Exception as e:
+        print(f"Error loading data: {e}")
+        return pd.DataFrame()
+
+def format_metadata(metadata_str):
+    """Format metadata as a readable string"""
+    try:
+        if isinstance(metadata_str, str):
+            metadata = json.loads(metadata_str)
+        else:
+            metadata = metadata_str
+        
+        formatted = []
+        for key, value in metadata.items():
+            if key == 'generated_at':
+                # Format timestamp
+                try:
+                    dt = datetime.fromisoformat(value.replace('Z', '+00:00'))
+                    formatted.append(f"**Generated:** {dt.strftime('%Y-%m-%d %H:%M UTC')}")
+                except:
+                    formatted.append(f"**Generated:** {value}")
+            else:
+                formatted.append(f"**{key.replace('_', ' ').title()}:** {value}")
+        
+        return "<br>".join(formatted)
+    except:
+        return str(metadata_str)
+
+def get_news_content(index):
+    """Get news content for a specific index"""
+    df = load_data()
+    
+    if df.empty or index >= len(df):
+        return (
+            "❌ No data available", 
+            "", 
+            f"0 / 0",
+            gr.update(interactive=False),
+            gr.update(interactive=False)
+        )
+    
+    row = df.iloc[index]
+    
+    # Format the main content
+    title = f"📰 HF News Summary - {row['summary_type'].title()} ({row['provider']})"
+    
+    # Create metadata table
+    metadata_html = f"""
+    <div style="background: #f8f9fa; padding: 15px; border-radius: 8px; margin: 10px 0;">
+        <h4>📊 Summary Details</h4>
+        <table style="width: 100%; font-size: 0.9em;">
+            <tr><td><strong>🗓️ Date Generated:</strong></td><td>{row['date_generated'].strftime('%Y-%m-%d %H:%M UTC')}</td></tr>
+            <tr><td><strong>⏰ Time Range:</strong></td><td>{row['time_range']}</td></tr>
+            <tr><td><strong>📈 Articles Analyzed:</strong></td><td>{row['num_articles_analyzed']}</td></tr>
+            <tr><td><strong>🔗 Sources:</strong></td><td>{row['sources']}</td></tr>
+            <tr><td><strong>🤖 Provider:</strong></td><td>{row['provider']}</td></tr>
+        </table>
+        <div style="margin-top: 10px;">
+            <strong>⚙️ Metadata:</strong><br>
+            {format_metadata(row['metadata'])}
+        </div>
+    </div>
+    """
+    
+    # Navigation info
+    nav_info = f"{index + 1} / {len(df)}"
+    
+    # Navigation buttons state
+    prev_disabled = index >= len(df) - 1
+    next_disabled = index <= 0
+    
+    return (
+        title,
+        metadata_html + "\n\n" + row['summarized_markdown'],
+        nav_info,
+        gr.update(interactive=not prev_disabled),
+        gr.update(interactive=not next_disabled)
+    )
+
+# Global state for current index
+current_index = 0
+
+def next_news():
+    global current_index
+    current_index = max(0, current_index - 1)  # Newer news (lower index)
+    return get_news_content(current_index)
+
+def prev_news():
+    global current_index
+    df = load_data()
+    current_index = min(len(df) - 1, current_index + 1)  # Older news (higher index)
+    return get_news_content(current_index)
+
+def refresh_data():
+    global current_index
+    current_index = 0  # Reset to latest
+    return get_news_content(current_index)
+
+# Create the Gradio interface
+with gr.Blocks(
+    title="HF News (AI-assisted summary)",
+    theme=gr.themes.Soft(),
+    css="""
+    .news-container {
+        max-width: 1200px;
+        margin: 0 auto;
+    }
+    .nav-button {
+        font-size: 16px !important;
+        padding: 10px 20px !important;
+    }
+    """
+) as app:
+    gr.HTML("""
+    <div style="text-align: center; padding: 20px;">
+        <h1>🚀 HF News (AI-assisted summary) 📰</h1>
+        <p style="font-size: 18px; color: #666;">
+            Stay updated with the latest AI and ML news, summarized by AI! ✨
+        </p>
+    </div>
+    """)
+    
+    with gr.Row():
+        with gr.Column(scale=1):
+            refresh_btn = gr.Button("🔄 Refresh Data", variant="secondary")
+        with gr.Column(scale=2):
+            nav_info = gr.Textbox(
+                label="📍 Position", 
+                interactive=False,
+                container=False
+            )
+        with gr.Column(scale=1):
+            pass  # Empty column for spacing
+    
+    title_display = gr.Textbox(
+        label="📋 Current Summary",
+        interactive=False,
+        container=False
+    )
+    
+    content_display = gr.Markdown(
+        label="📄 Content",
+        value="Loading news data... 🔄"
+    )
+    
+    with gr.Row():
+        prev_btn = gr.Button("⬅️ Older News", variant="secondary", elem_classes="nav-button")
+        next_btn = gr.Button("➡️ Newer News", variant="primary", elem_classes="nav-button")
+    
+    gr.HTML("""
+    <div style="text-align: center; padding: 20px; color: #666; font-size: 14px;">
+        <p>🤖 Powered by AI • 📊 Data from <a href="https://huggingface.co/datasets/rajatarya/hf-news-aggregator" target="_blank">HF News Aggregator</a></p>
+    </div>
+    """)
+    
+    # Event handlers
+    app.load(
+        fn=lambda: get_news_content(0),
+        outputs=[title_display, content_display, nav_info, prev_btn, next_btn]
+    )
+    
+    next_btn.click(
+        fn=next_news,
+        outputs=[title_display, content_display, nav_info, prev_btn, next_btn]
+    )
+    
+    prev_btn.click(
+        fn=prev_news,
+        outputs=[title_display, content_display, nav_info, prev_btn, next_btn]
+    )
+    
+    refresh_btn.click(
+        fn=refresh_data,
+        outputs=[title_display, content_display, nav_info, prev_btn, next_btn]
+    )
+
+if __name__ == "__main__":
+    app.launch() 

requirements.txt

@@ -0,0 +1,4 @@
+gradio>=4.0.0
+pandas>=1.5.0
+huggingface_hub>=0.16.0
+pyarrow>=10.0.0