Skip to main content
Version: 0.7.x

Interface: LLMType

Defined in: packages/react-native-executorch/src/types/llm.ts:35

React hook for managing a Large Language Model (LLM) instance.

Properties

configure()

configure: (configuration) => void

Defined in: packages/react-native-executorch/src/types/llm.ts:77

Configures chat and tool calling. See Configuring the model for details.

Parameters

configuration

LLMConfig

Configuration object containing chatConfig, toolsConfig, and generationConfig.

Returns

void


deleteMessage()

deleteMessage: (index) => void

Defined in: packages/react-native-executorch/src/types/llm.ts:120

Deletes all messages starting with message on index position. After deletion messageHistory will be updated.

Parameters

index

number

The index of the message to delete from history.

Returns

void


downloadProgress

downloadProgress: number

Defined in: packages/react-native-executorch/src/types/llm.ts:64

Represents the download progress as a value between 0 and 1, indicating the extent of the model file retrieval.


error

error: RnExecutorchError | null

Defined in: packages/react-native-executorch/src/types/llm.ts:69

Contains the error message if the model failed to load.


generate()

generate: (messages, tools?) => Promise<string>

Defined in: packages/react-native-executorch/src/types/llm.ts:92

Runs model to complete chat passed in messages argument. It doesn't manage conversation context.

Parameters

messages

Message[]

Array of messages representing the chat history.

tools?

Object[]

Optional array of tools that can be used during generation.

Returns

Promise<string>

The generated tokens as string.


getGeneratedTokenCount()

getGeneratedTokenCount: () => number

Defined in: packages/react-native-executorch/src/types/llm.ts:84

Returns the number of tokens generated so far in the current generation.

Returns

number

The count of generated tokens.


getPromptTokenCount()

getPromptTokenCount: () => number

Defined in: packages/react-native-executorch/src/types/llm.ts:104

Returns the number of prompt tokens in the last message.

Returns

number

The count of prompt token.


getTotalTokenCount()

getTotalTokenCount: () => number

Defined in: packages/react-native-executorch/src/types/llm.ts:98

Returns the number of total tokens from the previous generation.This is a sum of prompt tokens and generated tokens.

Returns

number

The count of prompt and generated tokens.


interrupt()

interrupt: () => void

Defined in: packages/react-native-executorch/src/types/llm.ts:125

Function to interrupt the current inference.

Returns

void


isGenerating

isGenerating: boolean

Defined in: packages/react-native-executorch/src/types/llm.ts:59

Indicates whether the model is currently generating a response.


isReady

isReady: boolean

Defined in: packages/react-native-executorch/src/types/llm.ts:54

Indicates whether the model is ready.


messageHistory

messageHistory: Message[]

Defined in: packages/react-native-executorch/src/types/llm.ts:39

History containing all messages in conversation. This field is updated after model responds to sendMessage.


response

response: string

Defined in: packages/react-native-executorch/src/types/llm.ts:44

State of the generated response. This field is updated with each token generated by the model.


sendMessage()

sendMessage: (message) => Promise<string>

Defined in: packages/react-native-executorch/src/types/llm.ts:113

Function to add user message to conversation. After model responds, messageHistory will be updated with both user message and model response.

Parameters

message

string

The message string to send.

Returns

Promise<string>

The model's response as a string.


token

token: string

Defined in: packages/react-native-executorch/src/types/llm.ts:49

The most recently generated token.