Summary

This PR adds support to run custom gemini and claude models using proxy urls

Explain the features implemented:

• A new backend type which is "proxy" has been introduced
• In model configuration if client type is specified then that will be used otherwise the model type will be used for creating the required client
• In custom models, classes have been added to run claude and gemini models using proxy urls

Performance impact (if any):

• N/A - New feature with no impact on existing functionality.

How to Test the feature

1. Run the glaive_code_assistant task using proxy models where client type and backend type are mentioned
2. Run the glaive_code_assistant task using any openai or vertexai models where client type is not mentioned

Screenshots (if applicable)

NA

Checklist

• Lint fixes and unit testing done
• End to end task testing
• Documentation updated