Init Astro + Tailwind + Shadcn-ui

- Added fallback model configuration in config.ini
- Updated requirements.txt to include rich for enhanced logging
- Refactored script.py to implement rich logging for better visibility
- Modified functions to include channel link extraction and updated CSV saving logic
- Improved error handling and user feedback throughout the script

.gitignore

@@ -0,0 +1,28 @@
+venv
+temp_thumbnail.png
+video_classifications.csv
+
+# build output
+dist/
+# generated types
+.astro/
+
+# dependencies
+node_modules/
+
+# logs
+npm-debug.log*
+yarn-debug.log*
+yarn-error.log*
+pnpm-debug.log*
+
+
+# environment variables
+.env
+.env.production
+
+# macOS-specific files
+.DS_Store
+
+# jetbrains setting folder
+.idea/

.vscode/extensions.json

@@ -0,0 +1,5 @@
+{
+  "recommendations": [
+    "astro-build.astro-vscode"
+  ]
+}

README.md

@@ -1,185 +1,214 @@
-# AutoDelete YouTube Videos 🗑️🎬
+# YouTube Video Classifier
 
-This script automates the process of deleting videos from your YouTube "Watch Later" playlist using PyAutoGUI.
+An AI-powered tool that automatically classifies YouTube videos in your "Watch Later" playlist based on their titles and thumbnails using vision-language models through Ollama.
 
----
+## Features ✨
 
-## Requirements 📦
+- 🤖 **AI-Powered Classification**: Uses Ollama with Qwen2.5-VL and fallback models to analyze video titles and thumbnails
+- 🔄 **Robust LLM Integration**: Automatic fallback between models with increasing timeouts for reliability
+- 📊 **Comprehensive CSV Storage**: Saves detailed video information including classifications, metadata, and thumbnails
+- 🌐 **Multi-language Detection**: Automatically detects video language using AI
+- 🏷️ **Smart Tagging**: Generates detailed sub-tags for better content organization
+- 🎯 **Smart Categories**: Uses existing classifications or creates new ones automatically
+- 🖥️ **Browser Automation**: Selenium-based interaction with YouTube for reliable data extraction
+- 🎨 **Beautiful Logging**: Rich console output with colors and emojis for better UX
+- ⌨️ **Easy Control**: Press 'q' at any time to safely quit the process
 
-- Python **3.11.10** 🐍
+## Quick Start
 
----
+### Prerequisites
+- Python 3.11.10+
+- Ollama installed locally
+- Chrome or Chromium browser
 
-## Setup & Usage (English) 🇬🇧
+### Setup
 
-### ⚙️ Requirements
+1. **Install Ollama**: Download from [https://ollama.ai](https://ollama.ai)
 
-- Python **3.11.10** 🐍
+2. **Pull Required Models**:
-
-### 1️⃣ Create a Virtual Environment
-
-**With `venv`:**
    ```bash
-python3.11 -m venv venv
+   ollama pull qwen2.5vl:7b
+   ollama pull gemma2:2b
    ```
 
-**With `virtualenv`:**
+3. **Start Ollama Service**:
    ```bash
-python3.11 -m pip install virtualenv
+   ollama serve
-python3.11 -m virtualenv venv
    ```
 
-### 2️⃣ Activate the Virtual Environment
+4. **Clone and Setup Project**:
-
-**On Linux/macOS:**
    ```bash
-source venv/bin/activate
+   git clone <repository-url>
-```
+   cd youtube-video-classifier
-**On Windows:**
-```bash
-venv\Scripts\activate
-```
    
-### 3️⃣ Install Dependencies
+   # Create virtual environment
+   python -m venv venv
+   source venv/bin/activate  # On Windows: venv\Scripts\activate
    
-```bash
+   # Install dependencies
    pip install -r requirements.txt
    ```
 
-### 4️⃣ Run the Script 🚀
+5. **Configure Settings** (optional):
+   Edit `config.ini` to customize your setup
 
+6. **Run the Classifier**:
    ```bash
    python script.py
    ```
 
----
+## How It Works 🔄
 
-## Features & Customization 🛠️
+1. **Browser Initialization**: Opens Chrome/Chromium and navigates to your YouTube "Watch Later" playlist
+2. **Video Detection**: Finds and extracts information from playlist videos using Selenium
+3. **Data Extraction**: Captures video title, thumbnail, channel info, duration, and upload date
+4. **AI Analysis**: Uses Ollama models to:
+   - Classify the video into categories
+   - Detect the primary language
+   - Generate detailed sub-tags
+5. **Smart Fallback**: If primary model fails/times out, automatically switches to fallback model
+6. **Data Storage**: Saves all information to CSV with base64-encoded thumbnails
+7. **Playlist Management**: Removes processed videos from "Watch Later" playlist
+8. **Continuous Processing**: Continues until all videos are processed or user quits
 
-### Features
+## Configuration
 
-- 🖼️ **Image Recognition:**  
+The `config.ini` file allows you to customize various settings:
-  Uses screenshots in the `img` folder to locate and interact with UI elements (browser icon, YouTube buttons, etc).
 
-- 🔍 **Region-based Search:**  
+```ini
-  The script splits the screen into left and right halves to speed up image searches.
+[DEFAULT]
+# Ollama settings
+ollama_host = http://localhost:11434
+ollama_model = qwen2.5vl:7b
+ollama_fallback_model = gemma2:2b
 
-- 🗂️ **Automatic Tab Handling:**  
+# File paths
-  Automatically opens/closes tabs and navigates to your "Watch Later" playlist.
+classifications_csv = video_classifications.csv
+playlist_url = https://www.youtube.com/playlist?list=WL
 
-- ⏱️ **Timing Controls:**  
+# LLM timeout settings (in seconds)
-  You can adjust `sleep` and `duration` values in the code to match your PC's speed.
+llm_primary_timeout = 60
+llm_fallback_timeout = 60
 
-- 🛑 **Hotkey Listener:**  
+# Processing settings
-  Press `q` at any time to safely stop the script.
+enable_delete = false
+enable_playlist_creation = false
+```
 
-- 🖱️ **Easy Customization:**  
+## CSV Output Format 📋
-  Change the images in the `img` folder or tweak the logic in functions like `locate_img` and `change_to_not_available` to adapt to UI changes or other platforms.
 
-### Customization
+The script creates a comprehensive CSV file with the following columns:
 
-- 🖼️ **Browser Image:**  
+- `video_title`: Title of the video
-  Replace the browser icon image in the `img` folder with a screenshot of your browser's icon. Make sure the filename matches the value of the `browser_img` variable in `script.py` (e.g., `brave.png` for Brave browser).
+- `video_url`: YouTube URL of the video
+- `thumbnail_url`: Path to the saved thumbnail
+- `classification`: AI-generated category
+- `language`: Detected language of the video
+- `channel_name`: Name of the YouTube channel
+- `channel_link`: URL to the channel
+- `video_length_seconds`: Duration in seconds
+- `video_date`: Upload date
+- `detailed_subtags`: AI-generated specific tags
+- `playlist_name`: Source playlist name
+- `playlist_link`: Source playlist URL
+- `image_data`: Base64-encoded thumbnail data
+- `timestamp`: When the classification was made
 
-- 🔗 **Playlist URL:**  
+## File Structure 📁
-  You can change the playlist URL by editing the value of the `playlist_url` variable at the top of the `script.py` file.
 
-- ⏱️ **Adjust Timing:**  
+```
-  The script uses `sleep` and `duration` values to wait for your PC to respond. You may need to increase or decrease these values depending on your computer's speed and internet connection.
+├── script.py                 # Main classification script
+├── config.ini               # Configuration settings
+├── requirements.txt         # Python dependencies
+├── video_classifications.csv # Generated results (created when first run)
+└── README.md               # This file
+```
 
-- 📌 **Pin Your Browser:**  
+## Features in Detail
-  For the script to work, your browser must be pinned to your taskbar.
+
+### AI Classification System
+- **Primary Model**: Qwen2.5-VL 7B for high-quality vision-language analysis
+- **Fallback Model**: Gemma2 2B for faster processing when primary model is slow
+- **Timeout Management**: Automatically increases timeout periods if models are struggling
+- **Continuous Retry**: Keeps trying until successful or user cancels
+
+### Data Extraction
+- **Video Metadata**: Title, URL, duration, upload date
+- **Channel Information**: Name and link to channel
+- **Thumbnail Capture**: Screenshots saved as base64 in CSV
+- **Playlist Context**: Source playlist name and URL
+
+### Browser Automation
+- **Multiple Chrome Paths**: Automatically finds Chrome/Chromium installation
+- **WebDriver Management**: Handles chromedriver setup and fallbacks
+- **Robust Selectors**: Multiple CSS selectors for reliable element finding
+- **Error Recovery**: Graceful handling of UI changes and loading delays
+
+### User Experience
+- **Rich Console Output**: Colored logging with emojis and status indicators
+- **Progress Tracking**: Clear indication of current processing status
+- **Safe Exit**: Press 'q' at any time to cleanly stop processing
+- **Error Reporting**: Detailed error messages for troubleshooting
+
+## Testing Your Setup
+
+Before running the main script, you can test individual components:
+
+1. **Test Ollama Connection**:
+   ```python
+   import requests
+   response = requests.get('http://localhost:11434/api/tags')
+   print(response.json())
+   ```
+
+2. **Test Browser Automation**:
+   Run the script and check if Chrome opens correctly
+
+3. **Test Model Response**:
+   The script will verify model availability on startup
+
+## Troubleshooting 🔧
+
+### Common Issues
+
+**Ollama Connection Error**:
+- Ensure Ollama is running: `ollama serve`
+- Check the host URL in config.ini
+- Verify models are installed: `ollama list`
+
+**Browser Issues**:
+- Install Chrome or Chromium
+- Update chromedriver if needed
+- Check if browser is in PATH
+
+**Model Timeout**:
+- The script automatically handles timeouts with fallback
+- Consider increasing timeout values in config.ini
+- Ensure sufficient system resources
+
+**Selenium Errors**:
+- YouTube may have changed their HTML structure
+- Check for browser updates
+- Verify you're logged into YouTube
+
+### Performance Tips
+
+- **For faster processing**: Use smaller models like `gemma2:2b` as primary
+- **For better accuracy**: Use larger models like `qwen2.5vl:7b` as primary
+- **For stability**: Keep both models installed for automatic fallback
+- **For large playlists**: Consider running in smaller batches
+
+## Contributing
+
+1. Fork the repository
+2. Create a feature branch
+3. Test your changes thoroughly
+4. Submit a pull request
+
+## License
+
+MIT License - see LICENSE file for details
 
 ---
 
-# AutoDelete YouTube Videos 🗑️🎬
+**Note**: This tool is for personal use and educational purposes. Please respect YouTube's Terms of Service and rate limits.
-
-Este script automatiza el proceso de eliminar videos de tu lista de "Ver más tarde" en YouTube usando PyAutoGUI.
-
----
-
-## Requisitos 📦
-
-- Python **3.11.10** 🐍
-
----
-
-## Configuración y Uso (Español) 🇪🇸
-
-### ⚙️ Requisitos
-
-- Python **3.11.10** 🐍
-
-### 1️⃣ Crear un Entorno Virtual
-
-**Con `venv`:**
-```bash
-python3.11 -m venv venv
-```
-
-**Con `virtualenv`:**
-```bash
-python3.11 -m pip install virtualenv
-python3.11 -m virtualenv venv
-```
-
-### 2️⃣ Activar el Entorno Virtual
-
-**En Linux/macOS:**
-```bash
-source venv/bin/activate
-```
-**En Windows:**
-```bash
-venv\Scripts\activate
-```
-
-### 3️⃣ Instalar las Dependencias
-
-```bash
-pip install -r requirements.txt
-```
-
-### 4️⃣ Ejecutar el Script 🚀
-
-```bash
-python script.py
-```
-
----
-
-## Funcionalidades y Personalización 🛠️
-
-### Funcionalidades
-
-- 🖼️ **Reconocimiento de Imágenes:**  
-  Usa capturas en la carpeta `img` para localizar e interactuar con elementos de la interfaz (icono del navegador, botones de YouTube, etc).
-
-- 🔍 **Búsqueda por Regiones:**  
-  El script divide la pantalla en mitades izquierda y derecha para acelerar la búsqueda de imágenes.
-
-- 🗂️ **Manejo Automático de Pestañas:**  
-  Abre/cierra pestañas y navega automáticamente a tu lista de "Ver más tarde".
-
-- ⏱️ **Control de Tiempos:**  
-  Puedes ajustar los valores de `sleep` y `duration` en el código según la velocidad de tu PC.
-
-- 🛑 **Escucha de Teclas:**  
-  Presiona `q` en cualquier momento para detener el script de forma segura.
-
-- 🖱️ **Fácil Personalización:**  
-  Cambia las imágenes en la carpeta `img` o ajusta la lógica en funciones como `locate_img` y `change_to_not_available` para adaptarlo a cambios en la interfaz o a otras plataformas.
-
-### Personalización
-
-- 🖼️ **Imagen del Navegador:**  
-  Reemplaza la imagen del icono de tu navegador en la carpeta `img` con una captura de pantalla del icono de tu navegador. Asegúrate de que el nombre del archivo coincida con el valor de la variable `browser_img` en `script.py` (por ejemplo, `brave.png` para el navegador Brave).
-
-- 🔗 **URL de la Playlist:**  
-  Puedes cambiar la URL de la playlist que se usa modificando el valor de la variable `playlist_url` al inicio del archivo `script.py`.
-
-- ⏱️ **Ajusta los Tiempos:**  
-  El script utiliza valores de `sleep` y `duration` para esperar la respuesta de tu PC. Puede que necesites aumentar o disminuir estos valores dependiendo de la velocidad de tu computadora y conexión a internet.
-
-- 📌 **Ancla tu Navegador:**  
-  Para que el script funcione, es indispensable que tengas tu navegador anclado a tu barra de tareas.

config.ini

@@ -0,0 +1,30 @@
+# Configuration file for YouTube Video Classifier
+
+[DEFAULT]
+# Ollama settings
+ollama_host = http://localhost:11434
+ollama_model = qwen2.5vl:7b
+ollama_fallback_model = gemma2:2b
+
+# File paths
+classifications_csv = video_classifications.csv
+browser_image = brave.png
+
+# YouTube settings
+playlist_url = https://www.youtube.com/playlist?list=WL
+
+# Image recognition settings
+confidence_threshold = 0.8
+search_timeout = 0.5
+sleep_duration = 0.2
+
+# Processing settings
+restart_tab_frequency = 90
+enable_delete = false
+enable_playlist_creation = false
+
+# LLM timeout settings (in seconds)
+llm_primary_timeout = 60
+llm_fallback_timeout = 60
+
+

demo_classification.py

@@ -0,0 +1,140 @@
+#!/usr/bin/env python3
+"""
+Demo script showing how the video classification works
+"""
+
+import requests
+import base64
+import time
+import configparser
+config = configparser.ConfigParser()
+config.read('config.ini')
+
+ollama_host = config.get('DEFAULT', 'ollama_host', fallback='http://ollama:11434')
+
+def classify_demo_video(video_obj):
+    """Demonstrate video classification."""
+    try:
+        # If there's a thumbnail, convert image to base64
+        if video_obj.get('thumbnail'):
+            with open(video_obj['thumbnail'], "rb") as image_file:
+                print(image_file.read())
+                image_data = base64.b64encode(image_file.read()).decode('utf-8')
+                print(image_data)
+        else:
+            image_data = None
+
+        existing_classifications = ["Tech Reviews", "Cooking", "Gaming", "Music"]
+        
+        prompt = f"""
+        Please classify this YouTube video based on its title and thumbnail.
+        
+        Video Title: {video_obj['title']}
+        
+        Existing Classifications: {", ".join(existing_classifications)}
+        
+        Instructions:
+        1. If the video fits into one of the existing classifications, use that exact classification name.
+        2. If the video doesn't fit any existing classification, create a new appropriate classification name.
+        3. Classification names should be concise (1-3 words) and descriptive.
+        4. Examples of good classifications: "Tech Reviews", "Cooking", "Gaming", "Education", "Music", "Comedy", etc.
+        5. Respond with ONLY the classification name, nothing else.
+        """
+        
+        print(f"Classifying: '{video_obj['title']}'")
+        print(f"Using thumbnail: {video_obj.get('thumbnail', 'None')}")
+        print("Sending request to Ollama...")
+        
+        # Prepare the request payload
+        payload = {
+            'model': 'qwen2.5vl:7b',
+            'prompt': prompt,
+            'stream': False
+        }
+        
+        # Only include images if image_data is available
+        if image_data:
+            payload['images'] = [image_data]
+        
+        response = requests.post(
+            f'{ollama_host}/api/generate',
+            json=payload,
+            timeout=60
+        )
+        
+        if response.status_code == 200:
+            result = response.json()
+            classification = result['response'].strip().strip('"\'')
+            print(f"✅ Classification: '{classification}'")
+            return classification
+        else:
+            print(f"❌ Error: {response.status_code}")
+            return "Uncategorized"
+            
+    except Exception as e:
+        print(f"❌ Error: {e}")
+        return "Uncategorized"
+
+def run_demo():
+    """Run classification demo with sample videos."""
+    sample_videos = [
+        {
+            "title": "I can't believe this change!", 
+            "thumbnail": "img/iphone_thumbnail.png"
+        },
+        {
+            "title": "iPhone 15 Pro Review - Best Camera Phone?"
+        },
+        {
+            "title": "Easy Pasta Recipe for Beginners", 
+        },
+        {
+            "title": "Minecraft Survival Guide - Episode 1", 
+        },
+        {
+            "title": "Classical Piano Music for Studying", 
+        },
+        {
+            "title": "Machine Learning Explained Simply", 
+        },
+    ]
+    
+    print("YouTube Video Classification Demo")
+    print("=" * 40)
+    
+    results = []
+    
+    for i, video_obj in enumerate(sample_videos, 1):
+        print(f"\n--- Demo {i}/{len(sample_videos)} ---")
+        
+        # Classify the video
+        classification = classify_demo_video(video_obj)
+        results.append((video_obj['title'], classification))
+        
+        time.sleep(1)  # Be nice to the API
+    
+    print("\n" + "=" * 40)
+    print("DEMO RESULTS:")
+    print("=" * 40)
+    
+    for title, classification in results:
+        print(f"{classification:15} | {title}")
+    
+    print("\nDemo complete! The script can:")
+    print("• Use existing categories when appropriate")
+    print("• Create new categories for unique content")
+    print("• Analyze both title and thumbnail information")
+
+if __name__ == '__main__':
+    print(ollama_host)
+    # Check if Ollama is running
+    try:
+        response = requests.get(f'{ollama_host}/api/tags', timeout=5)
+        if response.status_code != 200:
+            print("❌ Ollama is not running. Please start it with: ollama serve")
+            exit(1)
+    except:
+        print("❌ Cannot connect to Ollama. Please start it with: ollama serve")
+        exit(1)
+    
+    run_demo()

requirements.txt

@@ -2,4 +2,13 @@
 opencv-python==4.11.0.86
 pillow==11.3.0
 PyAutoGUI==0.9.54
-keyboard==0.13.5
+pynput==1.8.1
+requests==2.31.0
+pandas~=2.3.1
+ollama==0.2.1
+configparser==6.0.0
+pyperclip==1.8.2
+pytesseract==0.3.10
+selenium==4.15.2
+webdriver-manager==4.0.1
+rich==13.8.0

setup.sh

@@ -0,0 +1,63 @@
+#!/bin/bash
+
+# YouTube Video Classifier Setup Script
+
+echo "🎬 YouTube Video Classifier Setup"
+echo "=================================="
+
+# Check if Python 3.11 is available
+if ! command -v python3 &> /dev/null; then
+    echo "❌ Python 3.11 not found. Please install Python 3.11.10"
+    exit 1
+fi
+
+echo "✅ Python 3.11 found"
+
+# Create virtual environment
+echo "📦 Creating virtual environment..."
+python3 -m venv venv
+
+# Activate virtual environment
+echo "🔧 Activating virtual environment..."
+source venv/bin/activate
+
+# Install requirements
+echo "📥 Installing Python dependencies..."
+pip install -r requirements.txt
+
+# Check if Ollama is installed
+if ! command -v ollama &> /dev/null; then
+    echo "❌ Ollama not found. Please install Ollama from https://ollama.ai"
+    echo "   After installation, run:"
+    echo "   1. ollama serve"
+    echo "   2. ollama pull qwen2.5-vl:7b"
+    exit 1
+fi
+
+echo "✅ Ollama found"
+
+# Check if Ollama is running
+if ! curl -s http://localhost:11434/api/tags &> /dev/null; then
+    echo "⚠️  Ollama is not running. Starting Ollama..."
+    ollama serve &
+    sleep 5
+fi
+
+# Pull Qwen2.5VL model
+echo "🤖 Pulling Qwen2.5VL model..."
+ollama pull qwen2.5vl:7b
+
+# Test setup
+echo "🧪 Testing setup..."
+python test_ollama.py
+
+echo "✅ Setup complete!"
+echo ""
+echo "Next steps:"
+echo "1. Make sure your browser is pinned to the taskbar"
+echo "2. Update the browser image in img/ folder if needed"
+echo "3. Run: python script.py"
+echo ""
+echo "Optional:"
+echo "- Run demo: python demo_classification.py"
+echo "- Analyze results: python playlist_manager.py --analyze"

setup_model.py

@@ -0,0 +1,114 @@
+#!/usr/bin/env python3
+"""
+Script to ensure the Qwen2.5VL model is available in the Ollama container
+"""
+
+import configparser
+import time
+import sys
+import requests
+
+def load_config():
+    """Load configuration from config.ini"""
+    config = configparser.ConfigParser()
+    config.read('config.ini')
+
+    ollama_host = config.get('DEFAULT', 'ollama_host', fallback='http://ollama:11434')
+    ollama_model = config.get('DEFAULT', 'ollama_model', fallback='qwen2.5vl:7b')
+
+    return ollama_host, ollama_model
+
+def wait_for_ollama(host, max_attempts=30):
+    """Wait for Ollama container to be ready"""
+    print(f"⏳ Waiting for Ollama container at {host}...")
+
+    for attempt in range(1, max_attempts + 1):
+        try:
+            response = requests.get(f"{host}/api/tags", timeout=5)
+            if response.status_code == 200:
+                print("✅ Ollama container is ready!")
+                return True
+        except requests.exceptions.RequestException:
+            pass
+
+        print(f"   Attempt {attempt}/{max_attempts} - waiting...")
+        time.sleep(2)
+
+    print("❌ Ollama container is not responding after maximum attempts")
+    return False
+
+def check_model_exists(host, model_name):
+    """Check if the model is already available"""
+    try:
+        response = requests.get(f"{host}/api/tags", timeout=5)
+        if response.status_code == 200:
+            models = response.json()
+            model_names = [model['name'] for model in models.get('models', [])]
+            return any(model_name in name for name in model_names), model_names
+        return False, []
+    except requests.exceptions.RequestException as e:
+        print(f"❌ Error checking models: {e}")
+        return False, []
+
+def pull_model(host, model_name):
+    """Pull the model from Ollama"""
+    print(f"📥 Pulling model '{model_name}' (this may take several minutes)...")
+
+    try:
+        response = requests.post(
+            f"{host}/api/pull",
+            json={"name": model_name},
+            timeout=600  # 10 minutes timeout
+        )
+
+        if response.status_code == 200:
+            print(f"✅ Successfully pulled model '{model_name}'")
+            return True
+        else:
+            print(f"❌ Failed to pull model: HTTP {response.status_code}")
+            print(f"Response: {response.text}")
+            return False
+    except requests.exceptions.RequestException as e:
+        print(f"❌ Error pulling model: {e}")
+        return False
+
+def main():
+    """Main function to set up the model"""
+    print("🔧 Model Setup for YouTube Video Classifier")
+    print("=" * 50)
+
+    # Load configuration
+    try:
+        ollama_host, ollama_model = load_config()
+        print("📋 Configuration:")
+        print(f"   Host: {ollama_host}")
+        print(f"   Model: {ollama_model}")
+        print()
+    except Exception as e:
+        print(f"❌ Failed to load configuration: {e}")
+        sys.exit(1)
+
+    # Wait for Ollama to be ready
+    if not wait_for_ollama(ollama_host):
+        sys.exit(1)
+
+    # Check if model exists
+    model_exists, available_models = check_model_exists(ollama_host, ollama_model)
+
+    if model_exists:
+        print(f"✅ Model '{ollama_model}' is already available!")
+    else:
+        print(f"📋 Available models: {available_models}")
+        print(f"❌ Model '{ollama_model}' not found")
+
+        if pull_model(ollama_host, ollama_model):
+            print(f"🎉 Model '{ollama_model}' is now ready for use!")
+        else:
+            print(f"❌ Failed to set up model '{ollama_model}'")
+            sys.exit(1)
+
+    print("\n🎬 YouTube Video Classifier is ready!")
+    print("🧪 Run 'python test_ollama.py' to verify the setup")
+
+if __name__ == "__main__":
+    main()

test_ollama.py

@@ -0,0 +1,99 @@
+#!/usr/bin/env python3
+"""
+Test script to verify Ollama connection and Qwen2.5-VL model
+"""
+
+import requests
+import configparser
+
+def load_config():
+    """Load configuration from config.ini"""
+    config = configparser.ConfigParser()
+    config.read('config.ini')
+
+    ollama_host = config.get('DEFAULT', 'ollama_host', fallback='http://ollama:11434')
+    ollama_model = config.get('DEFAULT', 'ollama_model', fallback='qwen2.5vl:7b')
+
+    return ollama_host, ollama_model
+
+def test_ollama_connection(host, model_name):
+    """Test if Ollama is running and accessible."""
+    try:
+        response = requests.get(f'{host}/api/tags', timeout=5)
+        if response.status_code == 200:
+            models = response.json()
+            print("✅ Ollama is running!")
+            model_names = [model['name'] for model in models.get('models', [])]
+            print(f"Available models: {model_names}")
+
+            # Check if the configured model is available
+            model_available = any(model_name in name for name in model_names)
+            if model_available:
+                print(f"✅ {model_name} model is available!")
+            else:
+                print(f"❌ {model_name} model not found. Available models: {model_names}")
+                print(f"Model may still be downloading. Check with: curl {host}/api/tags")
+
+            return True
+        else:
+            print(f"❌ Ollama responded with status code: {response.status_code}")
+            return False
+    except requests.exceptions.ConnectionError:
+        print("❌ Cannot connect to Ollama container. Is the ollama service running?")
+        print("💡 Try: docker-compose up -d ollama")
+        return False
+    except Exception as e:
+        print(f"❌ Error checking Ollama: {e}")
+        return False
+
+def test_classification(host, model_name):
+    """Test a simple classification without image."""
+    try:
+        response = requests.post(
+            f'{host}/api/generate',
+            json={
+                'model': model_name,
+                'prompt': 'Classify this video title into a category: "How to Cook Pasta - Italian Recipe Tutorial". Respond with only the category name.',
+                'stream': False
+            },
+            timeout=30
+        )
+
+        if response.status_code == 200:
+            result = response.json()
+            classification = result['response'].strip()
+            print(f"✅ Test classification successful: '{classification}'")
+            return True
+        else:
+            print(f"❌ Classification test failed: {response.status_code}")
+            print(f"Response: {response.text}")
+            return False
+    except Exception as e:
+        print(f"❌ Error testing classification: {e}")
+        return False
+
+if __name__ == '__main__':
+    print("Testing Ollama setup for YouTube Video Classifier...")
+    print("-" * 50)
+
+    # Load configuration
+    try:
+        ollama_host, ollama_model = load_config()
+        print(f"📋 Configuration:")
+        print(f"   Host: {ollama_host}")
+        print(f"   Model: {ollama_model}")
+        print()
+    except Exception as e:
+        print(f"❌ Failed to load configuration: {e}")
+        exit(1)
+
+    if test_ollama_connection(ollama_host, ollama_model):
+        print("\nTesting classification...")
+        test_classification(ollama_host, ollama_model)
+
+    print("\nSetup verification complete!")
+    print("\nIf all tests passed, you can run the main script with: python script.py")
+    print("If any tests failed, please:")
+    print("1. Make sure the ollama container is running: docker-compose up -d ollama")
+    print(f"2. Wait for the model to download: curl {ollama_host}/api/tags")
+    print("3. Check container logs: docker-compose logs ollama")

web/README.md

@@ -0,0 +1,13 @@
+# Astro with Tailwind
+
+```sh
+pnpm create astro@latest -- --template with-tailwindcss
+```
+
+[![Open in StackBlitz](https://developer.stackblitz.com/img/open_in_stackblitz.svg)](https://stackblitz.com/github/withastro/astro/tree/latest/examples/with-tailwindcss)
+[![Open with CodeSandbox](https://assets.codesandbox.io/github/button-edit-lime.svg)](https://codesandbox.io/p/sandbox/github/withastro/astro/tree/latest/examples/with-tailwindcss)
+[![Open in GitHub Codespaces](https://github.com/codespaces/badge.svg)](https://codespaces.new/withastro/astro?devcontainer_path=.devcontainer/with-tailwindcss/devcontainer.json)
+
+Astro comes with [Tailwind](https://tailwindcss.com) support out of the box. This example showcases how to style your Astro project with Tailwind.
+
+For complete setup instructions, please see our [Tailwind Integration Guide](https://docs.astro.build/en/guides/integrations-guide/tailwind).

web/astro.config.mjs

@@ -0,0 +1,14 @@
+// @ts-check
+import { defineConfig } from 'astro/config';
+import tailwindcss from '@tailwindcss/vite';
+
+import react from '@astrojs/react';
+
+// https://astro.build/config
+export default defineConfig({
+  vite: {
+      plugins: [tailwindcss()]
+	},
+
+  integrations: [react()]
+});

web/components.json

@@ -0,0 +1,21 @@
+{
+  "$schema": "https://ui.shadcn.com/schema.json",
+  "style": "new-york",
+  "rsc": false,
+  "tsx": true,
+  "tailwind": {
+    "config": "",
+    "css": "src/styles/global.css",
+    "baseColor": "neutral",
+    "cssVariables": true,
+    "prefix": ""
+  },
+  "aliases": {
+    "components": "@/components",
+    "utils": "@/lib/utils",
+    "ui": "@/components/ui",
+    "lib": "@/lib",
+    "hooks": "@/hooks"
+  },
+  "iconLibrary": "lucide"
+}

web/package.json

@@ -0,0 +1,32 @@
+{
+  "name": "web",
+  "type": "module",
+  "version": "0.0.1",
+  "scripts": {
+    "dev": "astro dev",
+    "build": "astro build",
+    "preview": "astro preview",
+    "astro": "astro"
+  },
+  "dependencies": {
+    "@astrojs/mdx": "^4.3.0",
+    "@astrojs/react": "^4.3.0",
+    "@radix-ui/react-slot": "^1.2.3",
+    "@tailwindcss/vite": "^4.1.3",
+    "@types/canvas-confetti": "^1.9.0",
+    "@types/react": "^19.1.8",
+    "@types/react-dom": "^19.1.6",
+    "astro": "^5.11.0",
+    "canvas-confetti": "^1.9.3",
+    "class-variance-authority": "^0.7.1",
+    "clsx": "^2.1.1",
+    "lucide-react": "^0.525.0",
+    "react": "^19.1.0",
+    "react-dom": "^19.1.0",
+    "tailwind-merge": "^3.3.1",
+    "tailwindcss": "^4.1.3"
+  },
+  "devDependencies": {
+    "tw-animate-css": "^1.3.5"
+  }
+}

web/public/favicon.svg

@@ -0,0 +1,9 @@
+<svg xmlns="http://www.w3.org/2000/svg" fill="none" viewBox="0 0 128 128">
+    <path d="M50.4 78.5a75.1 75.1 0 0 0-28.5 6.9l24.2-65.7c.7-2 1.9-3.2 3.4-3.2h29c1.5 0 2.7 1.2 3.4 3.2l24.2 65.7s-11.6-7-28.5-7L67 45.5c-.4-1.7-1.6-2.8-2.9-2.8-1.3 0-2.5 1.1-2.9 2.7L50.4 78.5Zm-1.1 28.2Zm-4.2-20.2c-2 6.6-.6 15.8 4.2 20.2a17.5 17.5 0 0 1 .2-.7 5.5 5.5 0 0 1 5.7-4.5c2.8.1 4.3 1.5 4.7 4.7.2 1.1.2 2.3.2 3.5v.4c0 2.7.7 5.2 2.2 7.4a13 13 0 0 0 5.7 4.9v-.3l-.2-.3c-1.8-5.6-.5-9.5 4.4-12.8l1.5-1a73 73 0 0 0 3.2-2.2 16 16 0 0 0 6.8-11.4c.3-2 .1-4-.6-6l-.8.6-1.6 1a37 37 0 0 1-22.4 2.7c-5-.7-9.7-2-13.2-6.2Z" />
+    <style>
+        path { fill: #000; }
+        @media (prefers-color-scheme: dark) {
+            path { fill: #FFF; }
+        }
+    </style>
+</svg>

web/src/components/Button.astro

@@ -0,0 +1,19 @@
+---
+// Click button, get confetti!
+// Styled by Tailwind :)
+---
+
+<button
+	class="appearance-none py-2 px-4 bg-purple-500 text-white font-semibold rounded-lg shadow-md hover:bg-purple-700 focus:outline-none focus:ring-2 focus:ring-purple-400 focus:ring-opacity-75"
+>
+	<slot />
+</button>
+
+<script>
+	import confetti from 'canvas-confetti';
+	const button = document.body.querySelector('button');
+
+	if (button) {
+		button.addEventListener('click', () => confetti());
+	}
+</script>

web/src/components/ui/button.tsx

@@ -0,0 +1,59 @@
+import * as React from "react"
+import { Slot } from "@radix-ui/react-slot"
+import { cva, type VariantProps } from "class-variance-authority"
+
+import { cn } from "@/lib/utils"
+
+const buttonVariants = cva(
+  "inline-flex items-center justify-center gap-2 whitespace-nowrap rounded-md text-sm font-medium transition-all disabled:pointer-events-none disabled:opacity-50 [&_svg]:pointer-events-none [&_svg:not([class*='size-'])]:size-4 shrink-0 [&_svg]:shrink-0 outline-none focus-visible:border-ring focus-visible:ring-ring/50 focus-visible:ring-[3px] aria-invalid:ring-destructive/20 dark:aria-invalid:ring-destructive/40 aria-invalid:border-destructive",
+  {
+    variants: {
+      variant: {
+        default:
+          "bg-primary text-primary-foreground shadow-xs hover:bg-primary/90",
+        destructive:
+          "bg-destructive text-white shadow-xs hover:bg-destructive/90 focus-visible:ring-destructive/20 dark:focus-visible:ring-destructive/40 dark:bg-destructive/60",
+        outline:
+          "border bg-background shadow-xs hover:bg-accent hover:text-accent-foreground dark:bg-input/30 dark:border-input dark:hover:bg-input/50",
+        secondary:
+          "bg-secondary text-secondary-foreground shadow-xs hover:bg-secondary/80",
+        ghost:
+          "hover:bg-accent hover:text-accent-foreground dark:hover:bg-accent/50",
+        link: "text-primary underline-offset-4 hover:underline",
+      },
+      size: {
+        default: "h-9 px-4 py-2 has-[>svg]:px-3",
+        sm: "h-8 rounded-md gap-1.5 px-3 has-[>svg]:px-2.5",
+        lg: "h-10 rounded-md px-6 has-[>svg]:px-4",
+        icon: "size-9",
+      },
+    },
+    defaultVariants: {
+      variant: "default",
+      size: "default",
+    },
+  }
+)
+
+function Button({
+  className,
+  variant,
+  size,
+  asChild = false,
+  ...props
+}: React.ComponentProps<"button"> &
+  VariantProps<typeof buttonVariants> & {
+    asChild?: boolean
+  }) {
+  const Comp = asChild ? Slot : "button"
+
+  return (
+    <Comp
+      data-slot="button"
+      className={cn(buttonVariants({ variant, size, className }))}
+      {...props}
+    />
+  )
+}
+
+export { Button, buttonVariants }

web/src/layouts/main.astro

@@ -0,0 +1,17 @@
+---
+import { Button } from '@/components/ui/button';
+import '../styles/global.css';
+const { content } = Astro.props;
+---
+
+<html lang="en">
+	<head>
+		<meta charset="utf-8" />
+		<meta name="viewport" content="width=device-width" />
+		<link rel="icon" type="image/svg+xml" href="/favicon.svg" />
+		<title>{content.title}</title>
+	</head>
+	<body>
+		<slot />
+	</body>
+</html>

web/src/lib/utils.ts

@@ -0,0 +1,6 @@
+import { clsx, type ClassValue } from "clsx"
+import { twMerge } from "tailwind-merge"
+
+export function cn(...inputs: ClassValue[]) {
+  return twMerge(clsx(inputs))
+}

web/src/pages/index.astro

@@ -0,0 +1,27 @@
+---
+import '../styles/global.css';
+// Component Imports
+import Button from '../components/Button.astro';
+import {Button as ShadcnButton} from '../components/ui/button.tsx';
+
+// Full Astro Component Syntax:
+// https://docs.astro.build/basics/astro-components/
+---
+
+<html lang="en">
+	<head>
+		<meta charset="utf-8" />
+		<meta name="viewport" content="width=device-width" />
+		<link rel="icon" type="image/svg+xml" href="/favicon.svg" />
+		<meta name="generator" content={Astro.generator} />
+		<title>Astro + TailwindCSS</title>
+	</head>
+
+	<body>
+		<div class="grid place-items-center h-screen content-center">
+			<Button>Tailwind Button in Astro!</Button>
+			<a href="/markdown-page" class="p-4 underline">Markdown is also supported...</a>
+			<ShadcnButton>Shadcn Button</ShadcnButton>
+		</div>
+	</body>
+</html>

web/src/pages/markdown-page.md

@@ -0,0 +1,16 @@
+---
+title: 'Markdown + Tailwind'
+layout: ../layouts/main.astro
+---
+
+<div class="grid place-items-center h-screen content-center">
+ <div class="py-2 px-4 bg-purple-500 text-white font-semibold rounded-lg shadow-md">
+  Tailwind classes also work in Markdown!
+ </div>
+ <a
+  href="/"
+  class="p-4 underline hover:text-purple-500 transition-colors ease-in-out duration-200"
+ >
+  Go home
+ </a>
+</div>

web/src/styles/global.css

@@ -0,0 +1,124 @@
+@import 'tailwindcss';
+@import "tw-animate-css";
+
+@custom-variant dark (&:is(.dark *));
+
+@theme inline {
+  --radius-sm: calc(var(--radius) - 4px);
+  --radius-md: calc(var(--radius) - 2px);
+  --radius-lg: var(--radius);
+  --radius-xl: calc(var(--radius) + 4px);
+  --color-background: var(--background);
+  --color-foreground: var(--foreground);
+  --color-card: var(--card);
+  --color-card-foreground: var(--card-foreground);
+  --color-popover: var(--popover);
+  --color-popover-foreground: var(--popover-foreground);
+  --color-primary: var(--primary);
+  --color-primary-foreground: var(--primary-foreground);
+  --color-secondary: var(--secondary);
+  --color-secondary-foreground: var(--secondary-foreground);
+  --color-muted: var(--muted);
+  --color-muted-foreground: var(--muted-foreground);
+  --color-accent: var(--accent);
+  --color-accent-foreground: var(--accent-foreground);
+  --color-destructive: var(--destructive);
+  --color-border: var(--border);
+  --color-input: var(--input);
+  --color-ring: var(--ring);
+  --color-chart-1: var(--chart-1);
+  --color-chart-2: var(--chart-2);
+  --color-chart-3: var(--chart-3);
+  --color-chart-4: var(--chart-4);
+  --color-chart-5: var(--chart-5);
+  --color-sidebar: var(--sidebar);
+  --color-sidebar-foreground: var(--sidebar-foreground);
+  --color-sidebar-primary: var(--sidebar-primary);
+  --color-sidebar-primary-foreground: var(--sidebar-primary-foreground);
+  --color-sidebar-accent: var(--sidebar-accent);
+  --color-sidebar-accent-foreground: var(--sidebar-accent-foreground);
+  --color-sidebar-border: var(--sidebar-border);
+  --color-sidebar-ring: var(--sidebar-ring);
+}
+
+:root {
+  --radius: 0.625rem;
+  --background: oklch(1 0 0);
+  --foreground: oklch(0.145 0 0);
+  --card: oklch(1 0 0);
+  --card-foreground: oklch(0.145 0 0);
+  --popover: oklch(1 0 0);
+  --popover-foreground: oklch(0.145 0 0);
+  --primary: oklch(0.205 0 0);
+  --primary-foreground: oklch(0.985 0 0);
+  --secondary: oklch(0.97 0 0);
+  --secondary-foreground: oklch(0.205 0 0);
+  --muted: oklch(0.97 0 0);
+  --muted-foreground: oklch(0.556 0 0);
+  --accent: oklch(0.97 0 0);
+  --accent-foreground: oklch(0.205 0 0);
+  --destructive: oklch(0.577 0.245 27.325);
+  --border: oklch(0.922 0 0);
+  --input: oklch(0.922 0 0);
+  --ring: oklch(0.708 0 0);
+  --chart-1: oklch(0.646 0.222 41.116);
+  --chart-2: oklch(0.6 0.118 184.704);
+  --chart-3: oklch(0.398 0.07 227.392);
+  --chart-4: oklch(0.828 0.189 84.429);
+  --chart-5: oklch(0.769 0.188 70.08);
+  --sidebar: oklch(0.985 0 0);
+  --sidebar-foreground: oklch(0.145 0 0);
+  --sidebar-primary: oklch(0.205 0 0);
+  --sidebar-primary-foreground: oklch(0.985 0 0);
+  --sidebar-accent: oklch(0.97 0 0);
+  --sidebar-accent-foreground: oklch(0.205 0 0);
+  --sidebar-border: oklch(0.922 0 0);
+  --sidebar-ring: oklch(0.708 0 0);
+}
+
+.dark {
+  --background: oklch(0.145 0 0);
+  --foreground: oklch(0.985 0 0);
+  --card: oklch(0.205 0 0);
+  --card-foreground: oklch(0.985 0 0);
+  --popover: oklch(0.205 0 0);
+  --popover-foreground: oklch(0.985 0 0);
+  --primary: oklch(0.922 0 0);
+  --primary-foreground: oklch(0.205 0 0);
+  --secondary: oklch(0.269 0 0);
+  --secondary-foreground: oklch(0.985 0 0);
+  --muted: oklch(0.269 0 0);
+  --muted-foreground: oklch(0.708 0 0);
+  --accent: oklch(0.269 0 0);
+  --accent-foreground: oklch(0.985 0 0);
+  --destructive: oklch(0.704 0.191 22.216);
+  --border: oklch(1 0 0 / 10%);
+  --input: oklch(1 0 0 / 15%);
+  --ring: oklch(0.556 0 0);
+  --chart-1: oklch(0.488 0.243 264.376);
+  --chart-2: oklch(0.696 0.17 162.48);
+  --chart-3: oklch(0.769 0.188 70.08);
+  --chart-4: oklch(0.627 0.265 303.9);
+  --chart-5: oklch(0.645 0.246 16.439);
+  --sidebar: oklch(0.205 0 0);
+  --sidebar-foreground: oklch(0.985 0 0);
+  --sidebar-primary: oklch(0.488 0.243 264.376);
+  --sidebar-primary-foreground: oklch(0.985 0 0);
+  --sidebar-accent: oklch(0.269 0 0);
+  --sidebar-accent-foreground: oklch(0.985 0 0);
+  --sidebar-border: oklch(1 0 0 / 10%);
+  --sidebar-ring: oklch(0.556 0 0);
+}
+
+@layer base {
+  * {
+    @apply border-border outline-ring/50;
+  }
+  body {
+    @apply bg-background text-foreground;
+  }
+}
+
+button {
+  cursor: pointer;
+}

web/tsconfig.json

@@ -0,0 +1,20 @@
+{
+  "extends": "astro/tsconfigs/strict",
+  "include": [
+    ".astro/types.d.ts",
+    "**/*"
+  ],
+  "exclude": [
+    "dist"
+  ],
+  "compilerOptions": {
+    "jsx": "react-jsx",
+    "jsxImportSource": "react",
+    "baseUrl": ".",
+    "paths": {
+      "@/*": [
+        "./src/*"
+      ]
+    }
+  }
+}