[Fix] Fix lambda node failures

[psriramsnc]

Summary (Bug Fix)

This PR fixes multiple issues in the LambdaNode execution path and LangGraph backend integration:

• LambdaNode was not passing node_config to func_to_execute, causing unexpected runtime errors.
• vLLM model config attribute mismatch (model_serving_name vs model_name) caused failing tests and incorrect initialization.
  This PR standardizes it and updates tests accordingly.

🔧 Impacted Components

• 🧠 LambdaNode._exec_wrapper — now correctly calls func_to_execute(node_config, state)
• 🔌 BackendFactory & LangGraphFactory — fixed API mismatch, now accept exec_wrapper only
• 🚀 LiteLLMVLLM — fixes incorrect attribute name leading to wrong model initialization & test failures

🧪 How to Test

1. Run unit tests:

   make test-verbose

2. Create a simple LambdaNode and run:

• Ensure the function receives both node_config and state
• Validate LangGraph runnable executes without errors

3. Initialize LiteLLMVLLM with and without model_serving_name:

• Confirm model_name is populated correctly
• Updated tests in test_litellm_vllm.py should pass

📸 Screenshots

N/A

✅ Checklist

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

Notes

This PR unblocks LambdaNode execution and standardizes backend runnable handling.
No external API changes expected for end users.

[amitsnow]

LGTM 🚀