Module: lib/openai
This module provides interfaces to OpenAI's various models.
Type Aliases
ValidChatModel
Ƭ ValidChatModel: "gpt-3.5-turbo" | "gpt-3.5-turbo-0125" | "gpt-3.5-turbo-0301" | "gpt-3.5-turbo-0613" | "gpt-3.5-turbo-1106" | "gpt-3.5-turbo-16k" | "gpt-3.5-turbo-16k-0613" | "gpt-4" | "gpt-4-0125-preview" | "gpt-4-0613" | "gpt-4-1106-preview" | "gpt-4-32k" | "gpt-4-32k-0613" | "gpt-4-turbo" | "gpt-4-turbo-2024-04-09" | "gpt-4-turbo-preview"
Defined in
ValidCompletionModel
Ƭ ValidCompletionModel: "davinci-002" | "babbage-002" | "gpt-3.5-turbo-instruct"
Defined in
Variables
tokenizer
• Const tokenizer: Object
Type declaration
| Name | Type |
|---|---|
decode | (tokens: number[]) => string |
encode | (text: string) => number[] |
Defined in
Functions
DalleImageGen
▸ DalleImageGen(numSamples, size): AsyncGenerator<Element, Element[], unknown>
Generates an image from a prompt using the DALL-E model.
See
https://platform.openai.com/docs/guides/images/introduction
Parameters
| Name | Type | Description |
|---|---|---|
numSamples | ImageGenPropsWithChildren | The number of images to generate. Defaults to 1. |
size | ComponentContext | The size of the image to generate. Defaults to 512x512. |
Returns
AsyncGenerator<Element, Element[], unknown>
URL(s) to the generated image, wrapped in Image component(s).
Defined in
OpenAI
▸ OpenAI(«destructured»): Node
An AI.JSX component that invokes an OpenAI Large Language Model.
Parameters
| Name | Type |
|---|---|
«destructured» | { children: Node ; chatModel?: ValidChatModel ; client?: OpenAI ; completionModel?: ValidCompletionModel } & ModelProps |
Returns
Defined in
OpenAIChatModel
▸ OpenAIChatModel(props, «destructured»): RenderableStream
Represents an OpenAI text chat model (e.g., gpt-4).
Parameters
| Name | Type |
|---|---|
props | ModelProps & { children: Node } & { model: ValidChatModel ; logitBias?: Record<string, number> } |
«destructured» | ComponentContext |
Returns
Defined in
OpenAICompletionModel
▸ OpenAICompletionModel(props, «destructured»): RenderableStream
Represents an OpenAI text completion model (e.g., text-davinci-003).
Parameters
| Name | Type |
|---|---|
props | ModelProps & { children: Node } & { model: ValidCompletionModel ; logitBias?: Record<string, number> } |
«destructured» | ComponentContext |
Returns
Defined in
tokenCountForConversationMessage
▸ tokenCountForConversationMessage(message, render): Promise<number>
Parameters
| Name | Type |
|---|---|
message | ConversationMessage |
render | <TIntermediate>(renderable: Renderable, opts?: RenderOpts<TIntermediate, string>) => RenderResult<TIntermediate, string><TIntermediate>(renderable: Renderable, opts: RenderOpts<TIntermediate, PartiallyRendered[]>) => RenderResult<TIntermediate, PartiallyRendered[]> |
Returns
Promise<number>