I used this code with langchain 0.2.8 and it worked fine. I have to move the code to a new environment where I am limited to using langchain 0.1.12. When I run this code in the new environment I keep getting an error that .init takes 1 positional argument but 2 were provided. The error gets thrown with this line -> propositioning_llm = self.llm.with_structured_output(Sentences).

Is there some sort of different syntax I need to use for langchain 0.1.12?

  

        system_string = """

            [THIS CONTAINS INSTRUCTIONS ABOUT WHAT THE SYSTEM IS SUPPOSED TO DO]

        """


        propositioning_prompt = ChatPromptTemplate([
            ("system", system_string),
            ("human", "Decompose the following: {input}"),
        ])

        class Sentences(BaseModel):
            sentences: list[str]

        propositioning_llm = self.llm.with_structured_output(Sentences)

        propositioning_chain = propositioning_prompt | propositioning_llm

        if len(paragraph.lstrip()) > 5:
            sentences = propositioning_chain.invoke(paragraph)
            returned_list = sentences.sentences

        return returned_list