🤖 GPT-2 Local Autocomplete

📋 About This Extension

This is the VS Code extension component of GPT-2 Local Copilot. It provides the frontend interface for AI-powered code completion while keeping all AI processing local to your machine.

Key Features

• 🎨 Ghost Text UI: Inline code suggestions as you type
• ⚡ Real-time Completion: Debounced requests with 300ms delay
• 🔄 Smart Caching: Reuses predictions for consecutive typing
• 🎯 Multi-line Support: Intelligent detection for functions and classes
• 📊 Status Bar Integration: Visual toggle for extension control
• 🔧 Auto Server Management: Automatically starts/stops Python backend

🛠️ Technical Details

Architecture

VS Code Extension (TypeScript)
├── Inline Completion Provider
├── Status Bar Integration
├── Process Management
└── HTTP Client (Axios)

Requirements

• VS Code: 1.80.0 or later
• Python Backend: Running on http://127.0.0.1:8000
• Dependencies: Axios for HTTP requests

Supported Languages

• ✅ Python (optimized)
• ✅ JavaScript/TypeScript
• ✅ General programming languages

🚀 Quick Start

Development Mode

# Clone the repository
git clone https://github.com/prathmeshmagar447/gpt2-local.git
cd gpt2-local/extension

# Install dependencies
npm install

# Compile TypeScript
npm run compile

# Open in VS Code and press F5
code .

Using the Extension

1. Start Backend: Ensure Python server is running
2. Launch Extension: Press F5 in VS Code
3. Start Coding: Open any code file and begin typing
4. Accept Suggestions: Press Tab to accept ghost text

⚙️ Configuration

Extension Settings

Access via VS Code Settings (Ctrl/Cmd + ,):

{
  "codeCompletion.serverUrl": "http://127.0.0.1:8000"
}

Status Bar

• $(zap) GPT-2: ON: Extension active
• $(circle-slash) GPT-2: OFF: Extension disabled
• Click to toggle on/off

🔧 Development

Project Structure

extension/
├── src/
│   └── extension.ts      # Main extension logic
├── out/                  # Compiled JavaScript (generated)
├── package.json          # Extension manifest
├── tsconfig.json         # TypeScript configuration
└── icon.svg              # Extension icon

Available Scripts

npm run compile      # Build TypeScript to JavaScript
npm run watch        # Watch mode compilation
npm run pretest      # Run linting before tests
npm run lint         # ESLint code checking

Building for Distribution

# Install VS Code Extension CLI
npm install -g @vscode/vsce

# Package extension
vsce package

# Publish to marketplace
vsce publish

🔌 API Integration

The extension communicates with the Python backend via HTTP:

Request Format

POST /predict
{
  "code_context": "def hello",
  "multiline": false
}

Response Format

{
  "completion": "_world():\n    return 'Hello World'"
}

Error Handling

• Automatic retry on connection failures
• Graceful degradation when server is unavailable
• Detailed logging in VS Code Output panel

🐛 Troubleshooting

Extension Issues

Debug Information

• Logs: View → Output → Log (Extension Host)
• Server Status: Check http://127.0.0.1:8000/docs
• Network: Test with curl: curl -X POST http://127.0.0.1:8000/predict -H "Content-Type: application/json" -d '{"code_context": "test", "multiline": false}'

📊 Performance

Response Times

• Cache Hit: ~10-50ms
• GPU Inference: ~50-200ms
• CPU Inference: ~200-800ms

Resource Usage

• Memory: ~50MB (extension only)
• Network: Localhost HTTP requests
• CPU: Minimal background processing

🤝 Contributing

Code Contributions

1. Fork the repository
2. Create a feature branch
3. Make your changes
4. Test thoroughly
5. Submit a pull request

Extension Development

• Follow TypeScript best practices
• Use ESLint configuration
• Test on multiple platforms
• Document new features

Reporting Issues

• Use GitHub Issues for bugs
• Include VS Code version and OS
• Provide steps to reproduce
• Attach relevant logs

📄 License

MIT License - See main repository for details.

🔗 Links

• Main Repository: https://github.com/prathmeshmagar447/gpt2-local
• Issues: https://github.com/prathmeshmagar447/gpt2-local/issues
• Documentation: https://github.com/prathmeshmagar447/gpt2-local#readme