langchain.js
    Preparing search index...

    Class ChatOllama

    Deprecated in favor of the @langchain/ollama package. Import from @langchain/ollama instead.

    A class that enables calls to the Ollama API to access large language models in a chat-like fashion. It extends the SimpleChatModel class and implements the OllamaInput interface.

    const prompt = ChatPromptTemplate.fromMessages([
      [
        "system",
        `You are an expert translator. Format all responses as JSON objects with two keys: "original" and "translated".`,
      ],
      ["human", `Translate "{input}" into {language}.`],
    ]);

    const model = new ChatOllama({
      baseUrl: "http://api.example.com",
      model: "llama2",
      format: "json",
    });

    const chain = prompt.pipe(model);

    const result = await chain.invoke({
      input: "I love programming",
      language: "German",
    });

    Hierarchy (View Summary)

    Implements

    Index

    Constructors

    constructor

    Properties

    baseUrl embeddingOnly? f16KV? format? frequencyPenalty? headers? keepAlive lc_serializable logitsAll? lowVram? mainGpu? mirostat? mirostatEta? mirostatTau? model numBatch? numCtx? numGpu? numGqa? numKeep? numPredict? numThread? penalizeNewline? presencePenalty? repeatLastN? repeatPenalty? ropeFrequencyBase? ropeFrequencyScale? stop? temperature? tfsZ? topK? topP? typicalP? useMLock? useMMap? vocabOnly?

    Methods

    _combineLLMOutput _convertMessagesToOllamaMessages _formatMessagesAsPrompt _llmType _streamResponseChunks _streamResponseChunksLegacy getLsParams invocationParams lc_name

    Constructors

    • Parameters

      • fields: any

      Returns ChatOllama

    Properties

    baseUrl: string = "http://localhost:11434"
    embeddingOnly?: boolean
    f16KV?: boolean
    format?: any
    frequencyPenalty?: number
    headers?: Record<string, string>
    keepAlive: string = "5m"
    lc_serializable: boolean = true
    logitsAll?: boolean
    lowVram?: boolean
    mainGpu?: number
    mirostat?: number
    mirostatEta?: number
    mirostatTau?: number
    model: string = "llama2"
    numBatch?: number
    numCtx?: number
    numGpu?: number
    numGqa?: number
    numKeep?: number
    numPredict?: number
    numThread?: number
    penalizeNewline?: boolean
    presencePenalty?: number
    repeatLastN?: number
    repeatPenalty?: number
    ropeFrequencyBase?: number
    ropeFrequencyScale?: number
    stop?: string[]
    temperature?: number
    tfsZ?: number
    topK?: number
    topP?: number
    typicalP?: number
    useMLock?: boolean
    useMMap?: boolean
    vocabOnly?: boolean

    Methods

    _combineLLMOutput

    • _combineLLMOutput(): {}

      Returns {}

    Protected_convertMessagesToOllamaMessages

    • _convertMessagesToOllamaMessages(messages: BaseMessage[]): OllamaMessage[]

      Parameters

      • messages: BaseMessage[]

      Returns OllamaMessage[]

    Protected_formatMessagesAsPrompt

    • _formatMessagesAsPrompt(messages: BaseMessage[]): string

      Parameters

      • messages: BaseMessage[]

      Returns string

    _llmType

    • _llmType(): string

      Returns string

    _streamResponseChunks

    • _streamResponseChunks(
          input: BaseMessage[],
          options: unknown,
          runManager?: any,
      ): AsyncGenerator<ChatGenerationChunk>

      Parameters

      • input: BaseMessage[]
      • options: unknown
      • OptionalrunManager: any

      Returns AsyncGenerator<ChatGenerationChunk>

    _streamResponseChunksLegacy

    • _streamResponseChunksLegacy(
          input: BaseMessage[],
          options: unknown,
          runManager?: any,
      ): AsyncGenerator<ChatGenerationChunk>

      Parameters

      • input: BaseMessage[]
      • options: unknown
      • OptionalrunManager: any

      Returns AsyncGenerator<ChatGenerationChunk>

    • Parameters

      • options: unknown

      Returns LangSmithParams

    • A method that returns the parameters for an Ollama API call. It includes model and options parameters.

      Parameters

      • Optionaloptions: unknown

        Optional parsed call options.

      Returns {
          format: any;
          keep_alive: string;
          model: string;
          options: {
              embedding_only: undefined | boolean;
              f16_kv: undefined | boolean;
              frequency_penalty: undefined | number;
              logits_all: undefined | boolean;
              low_vram: undefined | boolean;
              main_gpu: undefined | number;
              mirostat: undefined | number;
              mirostat_eta: undefined | number;
              mirostat_tau: undefined | number;
              num_batch: undefined | number;
              num_ctx: undefined | number;
              num_gpu: undefined | number;
              num_gqa: undefined | number;
              num_keep: undefined | number;
              num_predict: undefined | number;
              num_thread: undefined | number;
              penalize_newline: undefined | boolean;
              presence_penalty: undefined | number;
              repeat_last_n: undefined | number;
              repeat_penalty: undefined | number;
              rope_frequency_base: undefined | number;
              rope_frequency_scale: undefined | number;
              stop: any;
              temperature: undefined | number;
              tfs_z: undefined | number;
              top_k: undefined | number;
              top_p: undefined | number;
              typical_p: undefined | number;
              use_mlock: undefined | boolean;
              use_mmap: undefined | boolean;
              vocab_only: undefined | boolean;
          };
      }

      An object containing the parameters for an Ollama API call.

    • Returns string

    Settings

    Member Visibility

    On This Page

    Constructors
    Properties
    Methods