Remote llama-swap multi-model configuration example

[Riyavesuwala]

This PR adds a practical example and documentation showing how to use
multiple models hosted on a remote llama-swap server with the VS Code plugin.

The example demonstrates:

• Dynamic model selection by name
• Switching models without restarting services
• CPU-friendly workflows using remote inference

This addresses the question raised in the issue about best practices
for configuring VS Code with a remote llama-swap setup.