Type Alias DatasetConfigurationDTO

DatasetConfigurationDTO: {
    BM25_AVG_LEN?: number | null;
    BM25_B?: number | null;
    BM25_ENABLED?: boolean | null;
    BM25_K?: number | null;
    DISTANCE_METRIC?: DistanceMetric | null;
    EMBEDDING_BASE_URL?: string | null;
    EMBEDDING_MODEL_NAME?: string | null;
    EMBEDDING_QUERY_PREFIX?: string | null;
    EMBEDDING_SIZE?: number | null;
    FREQUENCY_PENALTY?: number | null;
    FULLTEXT_ENABLED?: boolean | null;
    INDEXED_ONLY?: boolean | null;
    LLM_BASE_URL?: string | null;
    LLM_DEFAULT_MODEL?: string | null;
    LOCKED?: boolean | null;
    MAX_LIMIT?: number | null;
    MAX_TOKENS?: number | null;
    MESSAGE_TO_QUERY_PROMPT?: string | null;
    N_RETRIEVALS_TO_INCLUDE?: number | null;
    PRESENCE_PENALTY?: number | null;
    RAG_PROMPT?: string | null;
    RERANKER_BASE_URL?: string | null;
    SEMANTIC_ENABLED?: boolean | null;
    STOP_TOKENS?: string[] | null;
    SYSTEM_PROMPT?: string | null;
    TEMPERATURE?: number | null;
    USE_MESSAGE_TO_QUERY_PROMPT?: boolean | null;
}

Lets you specify the configuration for a dataset

Type declaration

  • OptionalBM25_AVG_LEN?: number | null

    The average length of the chunks in the index for BM25

  • OptionalBM25_B?: number | null

    The BM25 B parameter

  • OptionalBM25_ENABLED?: boolean | null

    Whether to use BM25

  • OptionalBM25_K?: number | null

    The BM25 K parameter

  • OptionalDISTANCE_METRIC?: DistanceMetric | null
  • OptionalEMBEDDING_BASE_URL?: string | null

    The base URL for the embedding API

  • OptionalEMBEDDING_MODEL_NAME?: string | null

    The name of the embedding model to use

  • OptionalEMBEDDING_QUERY_PREFIX?: string | null

    The prefix to use for the embedding query

  • OptionalEMBEDDING_SIZE?: number | null

    The size of the embeddings

  • OptionalFREQUENCY_PENALTY?: number | null

    The frequency penalty to use

  • OptionalFULLTEXT_ENABLED?: boolean | null

    Whether to use fulltext search

  • OptionalINDEXED_ONLY?: boolean | null

    Whether to only use indexed chunks

  • OptionalLLM_BASE_URL?: string | null

    The base URL for the LLM API

  • OptionalLLM_DEFAULT_MODEL?: string | null

    The default model to use for the LLM

  • OptionalLOCKED?: boolean | null

    Whether the dataset is locked to prevent changes or deletion

  • OptionalMAX_LIMIT?: number | null

    The maximum limit for the number of chunks for counting

  • OptionalMAX_TOKENS?: number | null

    The maximum number of tokens to use in LLM Response

  • OptionalMESSAGE_TO_QUERY_PROMPT?: string | null

    The prompt to use for converting a message to a query

  • OptionalN_RETRIEVALS_TO_INCLUDE?: number | null

    The number of retrievals to include with the RAG model

  • OptionalPRESENCE_PENALTY?: number | null

    The presence penalty to use

  • OptionalRAG_PROMPT?: string | null

    The prompt to use for the RAG model

  • OptionalRERANKER_BASE_URL?: string | null

    The base URL for the reranker API

  • OptionalSEMANTIC_ENABLED?: boolean | null

    Whether to use semantic search

  • OptionalSTOP_TOKENS?: string[] | null

    The stop tokens to use

  • OptionalSYSTEM_PROMPT?: string | null

    The system prompt to use for the LLM

  • OptionalTEMPERATURE?: number | null

    The temperature to use

  • OptionalUSE_MESSAGE_TO_QUERY_PROMPT?: boolean | null

    Whether to use the message to query prompt

Settings

Member Visibility