TaskWeaver is not able to handle the multiple LLM selection

[BrajkishorePrajapati]

Describe the bug
I am trying to create a chain-lit UI with a drop-down filter for selecting the LLM. But even after selecting a different model, I am getting the response from the default Model. Can someone please help me define the correct taskweaver.json so that it will select the right model for the response?

Also what changes need to be made in app.py

And this issue is different from the previous one where task weaver was not able to run for different llm for planner and code interpreter.

[liqul]

We didn't support configurations from Chainlit, which is pretty much for demonstration purpose.