Schema

Modules concerning the schema of GLLM Inference modules.

Activity

Bases: BaseModel

Base schema for any activity.

Attributes:

Name	Type	Description
type	str	The type of activity being performed. Defaults to an empty string.

ActivityEvent

Bases: Event

Event schema for model-triggered activities (e.g. web search, MCP call, etc.).

Attributes:

Name	Type	Description
id	str	The ID of the activity event. Defaults to None.
value	dict[str, Any]	The value of the activity event.
level	EventLevel	The severity level of the activity event. Defaults to EventLevel.INFO.
type	Literal[ACTIVITY]	The type of the activity event. Defaults to EventType.ACTIVITY.
timestamp	datetime	The timestamp of the activity event. Defaults to the current timestamp.
metadata	dict[str, Any]	The metadata of the activity event. Defaults to an empty dictionary.

from_activity(id_=None, activity=None) classmethod

Create an activity event from an Activity object.

Parameters:

Name	Type	Description	Default
id_	str | None	The ID of the activity event. Defaults to None.	None
activity	Activity | None	The activity object to create the event from. Defaults to None, in which case the value will be an empty dictionary.	None

Returns:

Name	Type	Description
ActivityEvent	ActivityEvent	The activity event.

Attachment

Bases: BaseModel

Defines a file attachment schema.

Attributes:

Name	Type	Description
data	bytes	The content data of the file attachment.
filename	str	The filename of the file attachment.
mime_type	str	The mime type of the file attachment.
extension	str	The extension of the file attachment.
url	str | None	The URL of the file attachment. Defaults to None.
metadata	dict[str, Any]	The metadata of the file attachment. Defaults to an empty dictionary.

__repr__()

Return string representation of the Attachment.

Returns:

Name	Type	Description
str	str	The string representation of the Attachment.

__str__()

Return string representation of the Attachment.

Returns:

Name	Type	Description
str	str	The string representation of the Attachment.

from_base64(base64_data, filename=None, metadata=None) classmethod

Creates an Attachment from a base64 string.

Parameters:

Name	Type	Description	Default
base64_data	str	The base64 string of the file.	required
filename	str | None	The filename of the file. Defaults to None, in which case the filename will be derived from the mime type.	None
metadata	dict[str, Any] | None	The metadata of the file attachment. Defaults to None, in which case an empty dictionary will be used.	None

Returns:

Name	Type	Description
Attachment	Attachment	The instantiated Attachment.

from_bytes(bytes, filename=None, metadata=None) classmethod

Creates an Attachment from bytes.

Parameters:

Name	Type	Description	Default
bytes	bytes	The bytes of the file.	required
filename	str | None	The filename of the file. Defaults to None, in which case the filename will be derived from the extension.	None
metadata	dict[str, Any] | None	The metadata of the file attachment. Defaults to None, in which case an empty dictionary will be used.	None

Returns:

Name	Type	Description
Attachment	Attachment	The instantiated Attachment.

from_data_url(data_url, filename=None, metadata=None) classmethod

Creates an Attachment from a data URL (data:[mime/type];base64,[bytes]).

Parameters:

Name	Type	Description	Default
data_url	str	The data URL of the file.	required
filename	str | None	The filename of the file. Defaults to None, in which case the filename will be derived from the mime type.	None
metadata	dict[str, Any] | None	The metadata of the file attachment. Defaults to None, in which case an empty dictionary will be used.	None

Returns:

Name	Type	Description
Attachment	Attachment	The instantiated Attachment.

from_path(path, filename=None, metadata=None) classmethod

Creates an Attachment from a path.

Parameters:

Name	Type	Description	Default
path	str	The path to the file.	required
filename	str | None	The filename of the file. Defaults to None, in which case the filename will be derived from the path.	None
metadata	dict[str, Any] | None	The metadata of the file attachment. Defaults to None, in which case an empty dictionary will be used.	None

Returns:

Name	Type	Description
Attachment	Attachment	The instantiated Attachment.

from_url(url, filename=None, metadata=None) classmethod

Creates an Attachment from a URL.

Parameters:

Name	Type	Description	Default
url	str	The URL of the file.	required
filename	str | None	The filename of the file. Defaults to None, in which case the filename will be derived from the URL.	None
metadata	dict[str, Any] | None	The metadata of the file attachment. Defaults to None, in which case an empty dictionary will be used.	None

Returns:

Name	Type	Description
Attachment	Attachment	The instantiated Attachment.

write_to_file(path=None)

Writes the Attachment to a file.

Parameters:

Name	Type	Description	Default
path	str | None	The path to the file. Defaults to None, in which case the filename will be used as the path.	None

AttachmentType

Bases: StrEnum

Defines valid attachment types.

BatchStatus

Bases: StrEnum

Defines the status of a batch job.

CodeEvent

Bases: BlockBasedEvent

Event schema for model-generated code to be executed.

Attributes:

Name	Type	Description
id	str	The ID of the code event. Defaults to None.
value	str	The value of the code event. Defaults to an empty string.
level	EventLevel	The severity level of the code event. Defaults to EventLevel.INFO.
type	CodeEventType	The type of the code event. Defaults to EventType.CODE.
timestamp	datetime	The timestamp of the code event. Defaults to the current timestamp.
metadata	dict[str, Any]	The metadata of the code event. Defaults to an empty dictionary.

CodeExecResult

Bases: BaseModel

Defines a code execution result when a language model is configured to execute code.

Attributes:

Name	Type	Description
id	str	The ID of the code execution. Defaults to an empty string.
code	str	The executed code. Defaults to an empty string.
output	list[str | Attachment]	The output of the executed code. Defaults to an empty list.

EmitDataType

Bases: StrEnum

Defines valid data types for emitting events.

InputTokenDetails

Bases: BaseModel

Defines the input token details schema.

Attributes:

Name	Type	Description
cached_tokens	int	The number of cached tokens. Defaults to 0.
uncached_tokens	int	The number of uncached tokens. Defaults to 0.

__add__(other)

Add two InputTokenDetails objects together.

Parameters:

Name	Type	Description	Default
other	InputTokenDetails	The other InputTokenDetails object to add.	required

Returns:

Name	Type	Description
InputTokenDetails	InputTokenDetails	A new InputTokenDetails object with summed values.

JinjaEnvType

Bases: StrEnum

Enumeration for Jinja environment types.

LMEventType

Bases: StrEnum

Defines event types to be emitted by the LM invoker.

LMEventTypeSuffix

Bases: StrEnum

Defines suffixes for LM event types.

LMOutput(*, outputs=None, token_usage=None, duration=None, finish_details=None, response='', structured_output=None, tool_calls=None, reasoning=None, attachments=None, citations=None, code_exec_results=None, mcp_calls=None)

Bases: BaseModel

Defines the output of a language model.

Attributes:

Name	Type	Description
outputs	list[LMOutputItem]	The outputs of the language model in sequential order. Defaults to an empty list.
token_usage	TokenUsage | None	The token usage analytics, if requested. Defaults to None.
duration	float | None	The duration of the invocation in seconds, if requested. Defaults to None.
finish_details	dict[str, Any]	The details about how the generation finished, if requested. Defaults to an empty dictionary.
text	str	The first text response.
structured_output	dict[str, Any] | BaseModel | None	The first structured output.
texts	list[str]	The texts from the outputs.
structured_outputs	list[dict[str, Any] | BaseModel]	The structured outputs from the outputs.
attachments	list[Attachment]	The attachments from the outputs.
tool_calls	list[ToolCall]	The tool calls from the outputs.
thinkings	list[Reasoning]	The thinkings from the outputs.
citations	list[Chunk]	The citations from the outputs.
code_exec_results	list[CodeExecResult]	The code exec results from the outputs.
mcp_calls	list[MCPCall]	The MCP calls from the outputs.
response	str	Deprecated. Replaced by text.
reasoning	list[Reasoning]	Deprecated. Replaced by thinkings.

Initialize the LMOutput.

This constructor is created for backward compatibility with the legacy method to initialize the LMOutput. This constructor will be removed in v0.6.

Parameters:

Name	Type	Description	Default
outputs	list[LMOutputItem] | None	The output items. Defaults to an empty list.	None
token_usage	TokenUsage | None	The token usage analytics. Defaults to None.	None
duration	float | None	The duration of the invocation in seconds. Defaults to None.	None
finish_details	dict[str, Any] | None	The details about how the generation finished. Defaults to an empty dictionary.	None
response	str	The first text response. Defaults to an empty string.	''
structured_output	dict[str, Any] | BaseModel | None	The first structured output. Defaults to None.	None
tool_calls	list[ToolCall] | None	The tool calls. Defaults to None.	None
reasoning	list[Reasoning] | None	The thinkings. Defaults to None.	None
attachments	list[Attachment] | None	The attachments. Defaults to None.	None
citations	list[Chunk] | None	The citations. Defaults to None.	None
code_exec_results	list[CodeExecResult] | None	The code exec results. Defaults to None.	None
mcp_calls	list[MCPCall] | None	The MCP calls. Defaults to None.	None

attachments: list[Attachment] property writable

Get the attachments from the LMOutput.

Returns:

Type	Description
list[Attachment]	list[Attachment]: The attachments from the LMOutput.

citations: list[Chunk] property writable

Get the citations from the LMOutput.

Returns:

Type	Description
list[Chunk]	list[Chunk]: The citations from the LMOutput.

code_exec_results: list[CodeExecResult] property writable

Get the code exec results from the LMOutput.

Returns:

Type	Description
list[CodeExecResult]	list[CodeExecResult]: The code exec results from the LMOutput.

mcp_calls: list[MCPCall] property writable

Get the MCP calls from the LMOutput.

Returns:

Type	Description
list[MCPCall]	list[MCPCall]: The MCP calls from the LMOutput.

reasoning: list[Reasoning] property writable

Deprecated property to get the thinkings from the LMOutput.

Returns:

Type	Description
list[Reasoning]	list[Reasoning]: The thinkings from the LMOutput.

response: str property writable

Deprecated property to get the first text response from the LMOutput.

Returns:

Name	Type	Description
str	str	The first text response from the LMOutput.

structured_output: dict[str, Any] | BaseModel | None property writable

Deprecated property to get the first structured output from the LMOutput.

Returns:

Type	Description
dict[str, Any] | BaseModel | None	dict[str, Any] | BaseModel | None: The first structured output from the LMOutput.

structured_outputs: list[dict[str, Any] | BaseModel] property

Get the structured outputs from the LMOutput.

Returns:

Type	Description
list[dict[str, Any] | BaseModel]	list[dict[str, Any] | BaseModel]: The structured outputs from the LMOutput.

text: str property

Get the first text from the LMOutput.

Returns:

Name	Type	Description
str	str	The first text from the LMOutput.

texts: list[str] property

Get the texts from the LMOutput.

Returns:

Type	Description
list[str]	list[str]: The texts from the LMOutput.

thinkings: list[Reasoning] property

Get the thinkings from the LMOutput.

Returns:

Type	Description
list[Reasoning]	list[Reasoning]: The thinkings from the LMOutput.

tool_calls: list[ToolCall] property writable

Get the tool calls from the LMOutput.

Returns:

Type	Description
list[ToolCall]	list[ToolCall]: The tool calls from the LMOutput.

__repr__()

Return string representation of the LMOutput.

Returns:

Name	Type	Description
str	str	The string representation of the LMOutput.

__str__()

Return string representation of the LMOutput.

Returns:

Name	Type	Description
str	str	The string representation of the LMOutput.

add_attachment(attachment)

Add an attachment or a list of attachments to the LMOutput.

Parameters:

Name	Type	Description	Default
attachment	Attachment | list[Attachment]	The attachment or a list of attachments to add.	required

add_citation(citation)

Add a citation or a list of citations to the LMOutput.

Parameters:

Name	Type	Description	Default
citation	Chunk | list[Chunk]	The citation or a list of citations to add.	required

add_code_exec_result(code_exec_result)

Add a code exec result or a list of code exec results to the LMOutput.

Parameters:

Name	Type	Description	Default
code_exec_result	CodeExecResult | list[CodeExecResult]	The code exec result or a list of code exec results to add.	required

add_mcp_call(mcp_call)

Add an MCP call or a list of MCP calls to the LMOutput.

Parameters:

Name	Type	Description	Default
mcp_call	MCPCall | list[MCPCall]	The MCP call or a list of MCP calls to add.	required

add_structured(structured)

Add a structured output or a list of structured outputs to the LMOutput.

Parameters:

Name	Type	Description	Default
structured	dict[str, Any] | BaseModel | list[dict[str, Any] | BaseModel]	The structured output or a list of structured outputs to add.	required

add_text(text)

Add an output or a list of outputs to the LMOutput.

Parameters:

Name	Type	Description	Default
text	str | list[str]	The text or a list of texts to add.	required

add_thinking(thinking)

Add a thinking or a list of thoughts to the LMOutput.

Parameters:

Name	Type	Description	Default
thinking	Reasoning | list[Reasoning]	The thinking or a list of thoughts to add.	required

add_tool_call(tool_call)

Add a tool call or a list of tool calls to the LMOutput.

Parameters:

Name	Type	Description	Default
tool_call	ToolCall | list[ToolCall]	The tool call or a list of tool calls to add.	required

LMOutputItem

Bases: BaseModel

Defines the output item of a language model.

Attributes:

Name	Type	Description
type	str	The type of the output item.
output	LMOutputData	The output data of the output item.

__repr__()

Return string representation of the LMOutputItem.

Returns:

Name	Type	Description
str	str	The string representation of the LMOutputItem.

__str__()

Return string representation of the LMOutputItem.

Returns:

Name	Type	Description
str	str	The string representation of the LMOutputItem.

LMOutputType

Bases: StrEnum

Defines valid types for language model outputs.

MCPCall

Bases: BaseModel

Defines an MCP call.

Attributes:

Name	Type	Description
id	str	The ID of the MCP call. Defaults to an empty string.
server_name	str	The name of the MCP server. Defaults to an empty string.
tool_name	str	The name of the tool. Defaults to an empty string.
args	dict[str, Any]	The arguments of the tool. Defaults to an empty dictionary.
output	str | None	The output of the tool. Defaults to None.

MCPCallActivity

Bases: Activity

Schema for MCP tool call.

Attributes:

Name	Type	Description
type	Literal[MCP_CALL]	The type of activity being performed. Defaults to ActivityType.MCP_CALL.
server_name	str	The name of the MCP server.
tool_name	str	The name of the tool.
args	dict[str, str]	The arguments of the tool.

MCPListToolsActivity

Bases: Activity

Schema for listing tools in MCP.

Attributes:

Name	Type	Description
type	Literal[MCP_LIST_TOOLS]	The type of activity being performed. Defaults to ActivityType.MCP_LIST_TOOLS.
server_name	str	The name of the MCP server. Defaults to an empty string.
tools	list[dict[str, str]] | None	The tools in the MCP server. Defaults to None.

MCPServer

Bases: BaseModel

Defines an MCP server.

Attributes:

Name	Type	Description
url	str	The URL of the MCP server.
name	str	The name of the MCP server.
allowed_tools	list[str] | None	The allowed tools of the MCP server. Defaults to None, in which case all tools are allowed.

Message

Bases: BaseModel

Defines a message schema to be used as inputs for a language model.

Attributes:

Name	Type	Description
role	MessageRole	The role of the message.
contents	list[MessageContent]	The contents of the message.
metadata	dict[str, Any]	The metadata of the message.

assistant(contents, metadata=None) classmethod

Create an assistant message.

Parameters:

Name	Type	Description	Default
contents	MessageContent | list[MessageContent]	The message contents. If a single content is provided, it will be wrapped in a list.	required
metadata	dict[str, Any]	Additional metadata for the message. Defaults to None.	None

Returns:

Name	Type	Description
Message	Message	A new message with ASSISTANT role.

system(contents, metadata=None) classmethod

Create a system message.

Parameters:

Name	Type	Description	Default
contents	MessageContent | list[MessageContent]	The message contents. If a single content is provided, it will be wrapped in a list.	required
metadata	dict[str, Any]	Additional metadata for the message. Defaults to None.	None

Returns:

Name	Type	Description
Message	Message	A new message with SYSTEM role.

user(contents, metadata=None) classmethod

Create a user message.

Parameters:

Name	Type	Description	Default
contents	MessageContent | list[MessageContent]	The message contents. If a single content is provided, it will be wrapped in a list.	required
metadata	dict[str, Any]	Additional metadata for the message. Defaults to None.	None

Returns:

Name	Type	Description
Message	Message	A new message with USER role.

MessageRole

Bases: StrEnum

Defines valid message roles.

ModelId

Bases: BaseModel

Defines a representation of a valid model id.

Attributes:

Name	Type	Description
provider	ModelProvider	The provider of the model.
name	str | None	The name of the model.
path	str | None	The path of the model.

Provider-specific examples

Using Anthropic

model_id = ModelId.from_string("anthropic/claude-sonnet-4-20250514")

Using Bedrock

model_id = ModelId.from_string("bedrock/us.anthropic.claude-sonnet-4-20250514-v1:0")

Using Cohere

model_id = ModelId.from_string("cohere/embed-english-v3.0")

Using Cohere with custom endpoint

model_id = ModelId.from_string("cohere/https://my-cohere-url:8000/v1:my-model-name")

Using Datasaur

model_id = ModelId.from_string("datasaur/https://deployment.datasaur.ai/api/deployment/teamId/deploymentId/")

Using Google

model_id = ModelId.from_string("google/gemini-2.5-flash-lite")

Using Jina

model_id = ModelId.from_string("jina/jina-embeddings-v2-large")

Using Jina with custom endpoint

model_id = ModelId.from_string("jina/https://my-jina-url:8000/v1:my-model-name")

Using OpenAI

model_id = ModelId.from_string("openai/gpt-5-nano")

Using OpenAI with Chat Completions API

model_id = ModelId.from_string("openai-chat-completions/gpt-5-nano")

Using OpenAI Responses API-compatible endpoints (e.g. SGLang)

model_id = ModelId.from_string("openai/https://my-sglang-url:8000/v1:my-model-name")

Using OpenAI Chat Completions API-compatible endpoints (e.g. Groq)

model_id = ModelId.from_string("openai-chat-completions/https://api.groq.com/openai/v1:llama3-8b-8192")

Using Azure OpenAI

model_id = ModelId.from_string("azure-openai/https://my-resource.openai.azure.com/openai/v1:my-deployment")

Using Voyage

model_id = ModelId.from_string("voyage/voyage-3.5-lite")

Using TwelveLabs

model_id = ModelId.from_string("twelvelabs/Marengo-retrieval-2.7")

Using LangChain

model_id = ModelId.from_string("langchain/langchain_openai.ChatOpenAI:gpt-4o-mini")

For the list of supported providers, please refer to the following table: https://python.langchain.com/docs/integrations/chat/#featured-providers

Using LiteLLM

model_id = ModelId.from_string("litellm/openai/gpt-4o-mini")

For the list of supported providers, please refer to the following page: https://docs.litellm.ai/docs/providers/

Using xAI

model_id = ModelId.from_string("xai/grok-4-0709")

For the list of supported models, please refer to the following page: https://docs.x.ai/docs/models

Custom model name validation example

validation_map = {
    ModelProvider.ANTHROPIC: {"claude-sonnet-4-20250514"},
    ModelProvider.GOOGLE: {"gemini-2.5-flash-lite"},
    ModelProvider.OPENAI: {"gpt-4.1-nano", "gpt-5-nano"},
}

model_id = ModelId.from_string("...", validation_map)

from_string(model_id, validation_map=None) classmethod

Parse a model id string into a ModelId object.

Parameters:

Name	Type	Description	Default
model_id	str	The model id to parse. Must be in the format defined in the following page: https://gdplabs.gitbook.io/sdk/resources/supported-models	required
validation_map	dict[str, set[str]] | None	An optional dictionary that maps provider names to sets of valid model names. For the defined model providers, the model names will be validated against the set of valid model names. For the undefined model providers, the model name will not be validated. Defaults to None.	None

Returns:

Name	Type	Description
ModelId	ModelId	The parsed ModelId object.

Raises:

Type	Description
ValueError	If the provided model id is invalid or if the model name is not valid for the provider.

to_string()

Convert the ModelId object to a string.

Returns:

Name	Type	Description
str	str	The string representation of the ModelId object. The format is defined in the following page: https://gdplabs.gitbook.io/sdk/resources/supported-models

ModelProvider

Bases: StrEnum

Defines the supported model providers.

OutputTokenDetails

Bases: BaseModel

Defines the output token details schema.

Attributes:

Name	Type	Description
reasoning_tokens	int	The number of reasoning tokens. Defaults to 0.
response_tokens	int	The number of response tokens. Defaults to 0.

__add__(other)

Add two OutputTokenDetails objects together.

Parameters:

Name	Type	Description	Default
other	OutputTokenDetails	The other OutputTokenDetails object to add.	required

Returns:

Name	Type	Description
OutputTokenDetails	OutputTokenDetails	A new OutputTokenDetails object with summed values.

Reasoning

Bases: BaseModel

Defines a reasoning output when a language model is configured to use reasoning.

Attributes:

Name	Type	Description
id	str	The ID of the reasoning output. Defaults to an empty string.
reasoning	str	The reasoning text. Defaults to an empty string.
type	str	The type of the reasoning output. Defaults to an empty string.
data	str	The additional data of the reasoning output. Defaults to an empty string.

ThinkingEvent

Bases: BlockBasedEvent

Event schema for model-generated thinking.

Attributes:

Name	Type	Description
id	str	The ID of the thinking event. Defaults to None.
value	str	The value of the thinking event. Defaults to an empty string.
level	EventLevel	The severity level of the thinking event. Defaults to EventLevel.INFO.
type	ThinkingEventType	The type of the thinking event. Defaults to EventType.THINKING.
timestamp	datetime	The timestamp of the thinking event. Defaults to the current timestamp.
metadata	dict[str, Any]	The metadata of the thinking event. Defaults to an empty dictionary.

TokenUsage

Bases: BaseModel

Defines the token usage data structure of a language model.

Attributes:

Name	Type	Description
input_tokens	int	The number of input tokens. Defaults to 0.
output_tokens	int	The number of output tokens. Defaults to 0.
input_token_details	InputTokenDetails | None	The details of the input tokens. Defaults to None.
output_token_details	OutputTokenDetails | None	The details of the output tokens. Defaults to None.

__add__(other)

Add two TokenUsage objects together.

Parameters:

Name	Type	Description	Default
other	TokenUsage	The other TokenUsage object to add.	required

Returns:

Name	Type	Description
TokenUsage	TokenUsage	A new TokenUsage object with summed values.

__repr__()

Return string representation of the TokenUsage.

Returns:

Name	Type	Description
str	str	The string representation of the TokenUsage.

__str__()

Return string representation of the TokenUsage.

Returns:

Name	Type	Description
str	str	The string representation of the TokenUsage.

from_token_details(input_tokens=None, output_tokens=None, cached_tokens=None, reasoning_tokens=None) classmethod

Creates a TokenUsage from token details.

Parameters:

Name	Type	Description	Default
input_tokens	int | None	The number of input tokens. Defaults to None.	None
output_tokens	int | None	The number of output tokens. Defaults to None.	None
cached_tokens	int | None	The number of cached tokens. Defaults to None.	None
reasoning_tokens	int | None	The number of reasoning tokens. Defaults to None.	None

Returns:

Name	Type	Description
TokenUsage	TokenUsage	The instantiated TokenUsage.

ToolCall

Bases: BaseModel

Defines a tool call request when a language model decides to invoke a tool.

Attributes:

Name	Type	Description
id	str	The ID of the tool call.
name	str	The name of the tool.
args	dict[str, Any]	The arguments of the tool call.

ToolResult

Bases: BaseModel

Defines a tool result to be sent back to the language model.

Attributes:

Name	Type	Description
id	str	The ID of the tool call.
output	str	The output of the tool call.

TruncateSide

Bases: StrEnum

Enumeration for truncation sides.

TruncationConfig

Bases: BaseModel

Configuration for text truncation behavior.

Attributes:

Name	Type	Description
max_length	int	Maximum length of text content. Required.
truncate_side	TruncateSide | None	Side to truncate from when max_length is exceeded. 1. TruncateSide.RIGHT: Keep the beginning of the text, truncate from the end (default) 2. TruncateSide.LEFT: Keep the end of the text, truncate from the beginning If None, defaults to TruncateSide.RIGHT

WebSearchActivity

Bases: Activity

Schema for web search tool call.

Attributes:

Name	Type	Description
type	WebSearchActivityTypes	The type of activity being performed. Defaults to ActivityType.SEARCH.
query	str | None	The query of the web search. Defaults to None.
url	str | None	The URL of the page. Defaults to None.
pattern	str | None	The pattern of the web search. Defaults to None.
sources	list[dict[str, str]] | None	The sources of the web search.

model_dump(*args, **kwargs)

Serialize the activity for display.

Returns:

Type	Description
dict[str, str]	dict[str, str]: The serialized activity.