feat: adding aws controller for llm application

[kkiani]

This pull request introduces several new examples and enhancements for working with LLMs (Large Language Models) using different approaches and tools. The changes include new example projects, additional functionality for AWS controllers, and improvements to existing components and resources.

New Example Projects:

• Dspy Application:

  • Added Pulumi.yaml configuration for a minimal Dspy application.
  • Implemented main application logic in __main__.py using AwsDspyController.
  • Defined a default question handler in applications.py.
  • Created AwsDspyController in controller.py to manage Dspy applications.
  • Updated requirements.txt with necessary dependencies.

• LLM with OpenAI Client:

  • Added Pulumi.yaml configuration for an OpenAI client application.
  • Implemented main application logic in __main__.py using AwsLlmController.
  • Updated requirements.txt with necessary dependencies.

• LLM with Simple Requests:

  • Added Pulumi.yaml configuration for a simple LLM application using open-source models.
  • Implemented main application logic in __main__.py using AwsLlmController and requests library.
  • Updated requirements.txt with necessary dependencies.

Enhancements to AWS Controllers:

• Base Controller Enhancements:

  • Added environment and is_runtime_execution properties to base_controller.py.

• LLM Controller:

  • Introduced LlmController as a base class for LLM-related controllers.
  • Created AwsLlmController to manage AWS-specific LLM operations, including SSM parameter retrieval and OpenAI client integration.

Improvements to AWS Resources:

• AwsVllmComponent:
  • Added support for storing endpoint URLs in SSM parameters. [1] [2]
  • Enhanced API Gateway resources to include versioning and specific routes for chat and completions. [1] [2] [3] [4] [5]