Add Web Automation Example with Structured Data Extraction

[jmanhype]

This PR adds a web automation example that demonstrates how to extract structured data from web content using instructor_ex.

Features

• Demonstrates integration with web automation services
• Shows how to extract structured data from web content
• Includes prompt engineering techniques for accurate extraction
• Implements error handling and timeout management
• NEW: Adds support for the proxy-lite-3b model from Convergence AI, a specialized 3B parameter Vision-Language Model designed for web automation

Usage

The example can be run with:

# Using the recommended proxy-lite-3b model:
./web_automation_example.exs "search query" --client proxy_lite_3b --homepage "https://en.wikipedia.org"

# Or using the original proxy service:
./web_automation_example.exs "search query" --client proxy_lite

What's included

This PR adds:

• web_automation_example.exs: The main example script
• web_automation_readme.md: Documentation for the example
• proxy_lite_example.py: Python script for using the proxy-lite-3b model

Benefits of proxy-lite-3b integration

• Purpose-built multimodal model designed specifically for web automation
• Makes decisions based on visual information from screenshots and HTML content
• More streamlined approach with fewer service dependencies
• Better handling of CAPTCHAs and other challenges when using Wikipedia as a starting point

Closes #93