[Question] RetryWithErrorOutputParser with_structured_output

[austinmw]

Hi, any idea if langchain's RetryWithErrorOutputParser is compatible with llm.with_structured_output()? Since LLMs can return incorrect responses that would cause Pydantic parsing to fail, it would be good to provide the previous output and error to the LLM when doing retries.

[adveatkarnik1]

Hi, did you get any clarity on this?

[austinmw]

Nope. Handled failures with LangGraph, but would be nice to have this built in.