from_llm() — langchain Function Reference

Architecture documentation for the from_llm() function in base.py from the langchain codebase.

Function python AgentOrchestration ClassicChains

Entity Profile

AgentOrchestration→ ClassicChains→ from_llm() — langchain Function Reference

Dependency Diagram

graph TD
  c53324f0_8b39_9054_fa8a_babe7e0eed82["from_llm()"]
  daa08d88_1d2b_5650_6257_d688074121b7["ElasticsearchDatabaseChain"]
  c53324f0_8b39_9054_fa8a_babe7e0eed82 -->|defined in| daa08d88_1d2b_5650_6257_d688074121b7
  style c53324f0_8b39_9054_fa8a_babe7e0eed82 fill:#6366f1,stroke:#818cf8,color:#fff

Relationship Graph

Source Code

libs/langchain/langchain_classic/chains/elasticsearch_database/base.py lines 177–208

    def from_llm(
        cls,
        llm: BaseLanguageModel,
        database: Elasticsearch,
        *,
        query_prompt: BasePromptTemplate | None = None,
        answer_prompt: BasePromptTemplate | None = None,
        query_output_parser: BaseOutputParser | None = None,
        **kwargs: Any,
    ) -> ElasticsearchDatabaseChain:
        """Convenience method to construct ElasticsearchDatabaseChain from an LLM.

        Args:
            llm: The language model to use.
            database: The Elasticsearch db.
            query_prompt: The prompt to use for query construction.
            answer_prompt: The prompt to use for answering user question given data.
            query_output_parser: The output parser to use for parsing model-generated
                ES query. Defaults to `SimpleJsonOutputParser`.
            kwargs: Additional arguments to pass to the constructor.
        """
        query_prompt = query_prompt or DSL_PROMPT
        query_output_parser = query_output_parser or SimpleJsonOutputParser()
        query_chain = query_prompt | llm | query_output_parser
        answer_prompt = answer_prompt or ANSWER_PROMPT
        answer_chain = answer_prompt | llm | StrOutputParser()
        return cls(
            query_chain=query_chain,
            answer_chain=answer_chain,
            database=database,
            **kwargs,
        )