langchain.js
    Preparing search index...

    Contains all relevant information returned by an LLM.

    type LLMResult = {
        __run?: Record<string, any>;
        generations: Generation[][];
        llmOutput?: Record<string, any>;
    }
    Index

    Properties

    __run?: Record<string, any>

    Dictionary of run metadata

    generations: Generation[][]

    List of the things generated. Each input could have multiple generations, hence this is a list of lists.

    llmOutput?: Record<string, any>

    Dictionary of arbitrary LLM-provider specific output.