Is Strict = True Default For Structured Outputs For Function Calling With Openai (Ex. FunctionCallingProgram)?

not at the moment (mostly because it has latency impacts, doesn't work with every pydantic class)

You can set it though

It will break for some pydantic classes though

since openai restricts it

I noticed that for some data models and prompts, when I tried to extract via FunctionCallingProgram I ran into an error (0 tool calls) but was able to get proper output using the response format (ie. not function calling) through openai - I'm assuming it's the structured output through function calling that is supported in llamaindex exclusively, correct? maybe if I set strict=True with the FunctionCallingProgram I will get the same output as calling OpenAI directly?

also, is there a simple way to set the system prompt for the FunctionCallingProgram? via chat_history or something else?

Pass in a ChatMessagePromptTemplate with your chat messages to FunctionCallingProgram

program(OutputCLS, prompt, arg1="val1", ...)

You can also set tool choice to force the function call

program(...., llm_kwargs={"tool_choice": "<class name>"})