Class ChatOpenAI<CallOptions>

Wrapper around OpenAI large language models that use the Chat endpoint.

To use you should have the openai package installed, with the OPENAI_API_KEY environment variable set.

To use with Azure you should have the openai package installed, with the AZURE_OPENAI_API_KEY, AZURE_OPENAI_API_INSTANCE_NAME, AZURE_OPENAI_API_DEPLOYMENT_NAME and AZURE_OPENAI_API_VERSION environment variable set. AZURE_OPENAI_BASE_PATH is optional and will override AZURE_OPENAI_API_INSTANCE_NAME if you need to use a custom endpoint.

Remarks

Any parameters that are valid to be passed to openai.createChatCompletion can be passed through modelKwargs, even if not explicitly available on this class.

Example

// Create a new instance of ChatOpenAI with specific temperature and model name settings
const model = new ChatOpenAI({
  temperature: 0.9,
  model: "ft:gpt-3.5-turbo-0613:{ORG_NAME}::{MODEL_ID}",
});

// Invoke the model with a message and await the response
const message = await model.invoke("Hi there!");

// Log the response to the console
console.log(message);

Type Parameters

Hierarchy (view full)

Implements

Constructors

constructor

Properties

frequencyPenalty model modelName n presencePenalty streaming temperature topP apiKey? azureOpenAIApiDeploymentName? azureOpenAIApiInstanceName? azureOpenAIApiKey? azureOpenAIApiVersion? azureOpenAIBasePath? logitBias? logprobs? maxTokens? modelKwargs? openAIApiKey? organization? stop? stopSequences? timeout? topLogprobs? user?

Accessors

callKeys

Methods

bindTools completionWithRetry getNumTokensFromMessages identifyingParams invocationParams withStructuredOutput

Constructors

constructor

Properties

frequencyPenalty

frequencyPenalty: number = 0

Penalizes repeated tokens according to frequency

model

model: string = "gpt-3.5-turbo"

Model name to use

modelName

modelName: string = "gpt-3.5-turbo"

Model name to use Alias for model

n

n: number = 1

Number of completions to generate for each prompt

presencePenalty

presencePenalty: number = 0

Penalizes repeated tokens

streaming

streaming: boolean = false

Whether to stream the results or not. Enabling disables tokenUsage reporting

temperature

temperature: number = 1

Sampling temperature to use

topP

topP: number = 1

Total probability mass of tokens to consider at each step

Optional apiKey

apiKey?: string

API key to use when making requests to OpenAI. Defaults to the value of OPENAI_API_KEY environment variable.

Optional azureOpenAIApiDeploymentName

azureOpenAIApiDeploymentName?: string

Azure OpenAI API deployment name to use for completions when making requests to Azure OpenAI. This is the name of the deployment you created in the Azure portal. e.g. "my-openai-deployment" this will be used in the endpoint URL: https://{InstanceName}.openai.azure.com/openai/deployments/my-openai-deployment/

Optional azureOpenAIApiInstanceName

azureOpenAIApiInstanceName?: string

Azure OpenAI API instance name to use when making requests to Azure OpenAI. this is the name of the instance you created in the Azure portal. e.g. "my-openai-instance" this will be used in the endpoint URL: https://my-openai-instance.openai.azure.com/openai/deployments/{DeploymentName}/

Optional azureOpenAIApiKey

azureOpenAIApiKey?: string

API key to use when making requests to Azure OpenAI.

Optional azureOpenAIApiVersion

azureOpenAIApiVersion?: string

API version to use when making requests to Azure OpenAI.

Optional azureOpenAIBasePath

azureOpenAIBasePath?: string

Custom endpoint for Azure OpenAI API. This is useful in case you have a deployment in another region. e.g. setting this value to "https://westeurope.api.cognitive.microsoft.com/openai/deployments" will be result in the endpoint URL: https://westeurope.api.cognitive.microsoft.com/openai/deployments/{DeploymentName}/

Optional logitBias

logitBias?: Record<string, number>

Dictionary used to adjust the probability of specific tokens being generated

Optional logprobs

logprobs?: boolean

Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message.

Optional maxTokens

maxTokens?: number

Maximum number of tokens to generate in the completion. -1 returns as many tokens as possible given the prompt and the model's maximum context size.

Optional modelKwargs

modelKwargs?: Record<string, any>

Holds any additional parameters that are valid to pass to openai.createCompletion that are not explicitly specified on this class.

Optional openAIApiKey

openAIApiKey?: string

API key to use when making requests to OpenAI. Defaults to the value of OPENAI_API_KEY environment variable. Alias for apiKey

Optional organization

organization?: string

Optional stop

stop?: string[]

List of stop words to use when generating Alias for stopSequences

Optional stopSequences

stopSequences?: string[]

List of stop words to use when generating

Optional timeout

timeout?: number

Timeout to use when making requests to OpenAI.

Optional topLogprobs

topLogprobs?: number

An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability. logprobs must be set to true if this parameter is used.

Optional user

user?: string

Unique string identifier representing your end-user, which can help OpenAI to monitor and detect abuse.

Accessors

callKeys

Methods

bindTools

completionWithRetry

  • completionWithRetry(request, options?): Promise<AsyncIterable<ChatCompletionChunk>>

  • Calls the OpenAI API with retry logic in case of failures.

    Parameters

    • request: ChatCompletionCreateParamsStreaming

      The request to send to the OpenAI API.

    • Optional options: OpenAICoreRequestOptions

      Optional configuration for the API call.

    Returns Promise<AsyncIterable<ChatCompletionChunk>>

    The response from the OpenAI API.

  • completionWithRetry(request, options?): Promise<ChatCompletion>

  • Parameters

    Returns Promise<ChatCompletion>

getNumTokensFromMessages

  • getNumTokensFromMessages(messages): Promise<{
        countPerMessage: any[];
        totalCount: number;
    }>

  • Parameters

    • messages: BaseMessage[]

    Returns Promise<{
        countPerMessage: any[];
        totalCount: number;
    }>

identifyingParams

  • identifyingParams(): Omit<ChatCompletionCreateParams, "messages"> & {
        model_name: string;
    } & ClientOptions

  • Get the identifying parameters for the model

    Returns Omit<ChatCompletionCreateParams, "messages"> & {
        model_name: string;
    } & ClientOptions

invocationParams

  • invocationParams(options?): Omit<ChatCompletionCreateParams, "messages">

  • Get the parameters used to invoke the model

    Parameters

    • Optional options: unknown

    Returns Omit<ChatCompletionCreateParams, "messages">

withStructuredOutput

  • withStructuredOutput<RunOutput>(outputSchema, config?): Runnable<BaseLanguageModelInput, RunOutput>

  • Type Parameters

    • RunOutput extends Record<string, any> = Record<string, any>

    Parameters

    • outputSchema: any
    • Optional config: any

    Returns Runnable<BaseLanguageModelInput, RunOutput>

  • withStructuredOutput<RunOutput>(outputSchema, config?): Runnable<BaseLanguageModelInput, {
        parsed: RunOutput;
        raw: BaseMessage;
    }>

  • Type Parameters

    • RunOutput extends Record<string, any> = Record<string, any>

    Parameters

    • outputSchema: any
    • Optional config: any

    Returns Runnable<BaseLanguageModelInput, {
        parsed: RunOutput;
        raw: BaseMessage;
    }>

Settings

Member Visibility

Theme

On This Page

Generated using TypeDoc