openai-core/com.aallam.openai.api.assistant/AssistantRequest

AssistantRequest

@Serializable

data class AssistantRequest( val model: ModelId? = null, val name: String? = null, val description: String? = null, val instructions: String? = null, val tools: List<AssistantTool>? = null, val toolResources: ToolResources? = null, val metadata: Map<String, String>? = null, val temperature: Double? = null, val topP: Double? = null, val responseFormat: AssistantResponseFormat? = null)

Constructors

AssistantRequest

constructor( model: ModelId? = null, name: String? = null, description: String? = null, instructions: String? = null, tools: List<AssistantTool>? = null, toolResources: ToolResources? = null, metadata: Map<String, String>? = null, temperature: Double? = null, topP: Double? = null, responseFormat: AssistantResponseFormat? = null)

Properties

description

@SerialName(value = "description")

val description: String? = null

The description of the assistant. Optional. The maximum length is 512 characters.

instructions

@SerialName(value = "instructions")

val instructions: String? = null

The system instructions that the assistant uses. Optional. The maximum length is 32768 characters.

metadata

@SerialName(value = "metadata")

val metadata: Map<String, String>? = null

Set of 16 key-value pairs that can be attached to an object. Optional. Keys can be a maximum of 64 characters long, and values can be a maximum of 512 characters long.

model

@SerialName(value = "model")

val model: ModelId? = null

ID of the model to use. This is required if the assistant does not yet exist.

name

@SerialName(value = "name")

val name: String? = null

The name of the assistant. Optional. The maximum length is 256 characters.

responseFormat

@SerialName(value = "response_format")

val responseFormat: AssistantResponseFormat? = null

Specifies the format that the model must output. Compatible with GPT-4o, GPT-4 Turbo, and all GPT-3.5 Turbo models since gpt-3.5-turbo-1106.

temperature

@SerialName(value = "temperature")

val temperature: Double? = null

What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.

toolResources

@SerialName(value = "tool_resources")

val toolResources: ToolResources? = null

A set of resources that are used by the assistant's tools. The resources are specific to the type of tool. For example, the code_interpreter tool requires a list of file IDs, while the file_search tool requires a list of vector store IDs.

tools

@SerialName(value = "tools")

val tools: List<AssistantTool>? = null

A list of tools enabled on the assistant. Optional. Defaults to an empty list. Tools can be of types code_interpreter, retrieval, or function.

topP

@SerialName(value = "top_p")

val topP: Double? = null

An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.