Skip to content

Annotators

zenml.annotators special

Initialization of the ZenML annotator stack component.

base_annotator

Base class for ZenML annotator stack components.

BaseAnnotator (StackComponent, ABC) pydantic-model

Base class for all ZenML annotators.

Attributes:

Name Type Description
notebook_only ClassVar[bool]

if the annotator can only be used in a notebook.
Source code in zenml/annotators/base_annotator.py 
class BaseAnnotator(StackComponent, ABC):
    """Base class for all ZenML annotators.

    Attributes:
        notebook_only: if the annotator can only be used in a notebook.
    """

    notebook_only: ClassVar[bool] = False

    # Class configuration
    TYPE: ClassVar[StackComponentType] = StackComponentType.ANNOTATOR
    FLAVOR: ClassVar[str]

    @abstractmethod
    def get_url(self) -> str:
        """Gets the URL of the annotation interface.

        Returns:
            The URL of the annotation interface.
        """

    @abstractmethod
    def get_url_for_dataset(self, dataset_name: str) -> str:
        """Gets the URL of the annotation interface for a specific dataset.

        Args:
            dataset_name: name of the dataset.

        Returns:
            The URL of the dataset annotation interface.
        """

    @abstractmethod
    def get_datasets(self) -> List[Any]:
        """Gets the datasets currently available for annotation.

        Returns:
            The datasets currently available for annotation.
        """

    @abstractmethod
    def get_dataset_names(self) -> List[str]:
        """Gets the names of the datasets currently available for annotation.

        Returns:
            The names of the datasets currently available for annotation.
        """

    @abstractmethod
    def get_dataset_stats(self, dataset_name: str) -> Tuple[int, int]:
        """Gets the statistics of a dataset.

        Args:
            dataset_name: name of the dataset.

        Returns:
            A tuple containing (labeled_task_count, unlabeled_task_count) for
                the dataset.
        """

    @abstractmethod
    def launch(self, url: Optional[str]) -> None:
        """Launches the annotation interface.

        Args:
            url: The URL of the annotation interface.
        """

    @abstractmethod
    def add_dataset(self, **kwargs: Any) -> Any:
        """Registers a dataset for annotation.

        Args:
            **kwargs: keyword arguments.

        Returns:
            The dataset or confirmation object on adding the dataset.
        """

    @abstractmethod
    def get_dataset(self, **kwargs: Any) -> Any:
        """Gets the dataset with the given name.

        Args:
            **kwargs: keyword arguments.

        Returns:
            The dataset with the given name.
        """

    @abstractmethod
    def delete_dataset(self, **kwargs: Any) -> None:
        """Deletes a dataset.

        Args:
            **kwargs: keyword arguments.
        """

    @abstractmethod
    def get_labeled_data(self, **kwargs: Any) -> Any:
        """Gets the labeled data for the given dataset.

        Args:
            **kwargs: keyword arguments.

        Returns:
            The labeled data for the given dataset.
        """

    @abstractmethod
    def get_unlabeled_data(self, **kwargs: str) -> Any:
        """Gets the unlabeled data for the given dataset.

        Args:
            **kwargs: Additional keyword arguments to pass to the Label Studio client.

        Returns:
            The unlabeled data for the given dataset.
        """
add_dataset(self, **kwargs)

Registers a dataset for annotation.

Parameters:

Name Type Description Default
**kwargs Any

keyword arguments.

 {}

Returns:

Type Description
Any

The dataset or confirmation object on adding the dataset.
Source code in zenml/annotators/base_annotator.py 
@abstractmethod
def add_dataset(self, **kwargs: Any) -> Any:
    """Registers a dataset for annotation.

    Args:
        **kwargs: keyword arguments.

    Returns:
        The dataset or confirmation object on adding the dataset.
    """
delete_dataset(self, **kwargs)

Deletes a dataset.

Parameters:

Name Type Description Default
**kwargs Any

keyword arguments.

 {}
Source code in zenml/annotators/base_annotator.py 
@abstractmethod
def delete_dataset(self, **kwargs: Any) -> None:
    """Deletes a dataset.

    Args:
        **kwargs: keyword arguments.
    """
get_dataset(self, **kwargs)

Gets the dataset with the given name.

Parameters:

Name Type Description Default
**kwargs Any

keyword arguments.

 {}

Returns:

Type Description
Any

The dataset with the given name.
Source code in zenml/annotators/base_annotator.py 
@abstractmethod
def get_dataset(self, **kwargs: Any) -> Any:
    """Gets the dataset with the given name.

    Args:
        **kwargs: keyword arguments.

    Returns:
        The dataset with the given name.
    """
get_dataset_names(self)

Gets the names of the datasets currently available for annotation.

Returns:

Type Description
List[str]

The names of the datasets currently available for annotation.
Source code in zenml/annotators/base_annotator.py 
@abstractmethod
def get_dataset_names(self) -> List[str]:
    """Gets the names of the datasets currently available for annotation.

    Returns:
        The names of the datasets currently available for annotation.
    """
get_dataset_stats(self, dataset_name)

Gets the statistics of a dataset.

Parameters:

Name Type Description Default
dataset_name str

name of the dataset.

 required

Returns:

Type Description
Tuple[int, int]

A tuple containing (labeled_task_count, unlabeled_task_count) for the dataset.
Source code in zenml/annotators/base_annotator.py 
@abstractmethod
def get_dataset_stats(self, dataset_name: str) -> Tuple[int, int]:
    """Gets the statistics of a dataset.

    Args:
        dataset_name: name of the dataset.

    Returns:
        A tuple containing (labeled_task_count, unlabeled_task_count) for
            the dataset.
    """
get_datasets(self)

Gets the datasets currently available for annotation.

Returns:

Type Description
List[Any]

The datasets currently available for annotation.
Source code in zenml/annotators/base_annotator.py 
@abstractmethod
def get_datasets(self) -> List[Any]:
    """Gets the datasets currently available for annotation.

    Returns:
        The datasets currently available for annotation.
    """
get_labeled_data(self, **kwargs)

Gets the labeled data for the given dataset.

Parameters:

Name Type Description Default
**kwargs Any

keyword arguments.

 {}

Returns:

Type Description
Any

The labeled data for the given dataset.
Source code in zenml/annotators/base_annotator.py 
@abstractmethod
def get_labeled_data(self, **kwargs: Any) -> Any:
    """Gets the labeled data for the given dataset.

    Args:
        **kwargs: keyword arguments.

    Returns:
        The labeled data for the given dataset.
    """
get_unlabeled_data(self, **kwargs)

Gets the unlabeled data for the given dataset.

Parameters:

Name Type Description Default
**kwargs str

Additional keyword arguments to pass to the Label Studio client.

 {}

Returns:

Type Description
Any

The unlabeled data for the given dataset.
Source code in zenml/annotators/base_annotator.py 
@abstractmethod
def get_unlabeled_data(self, **kwargs: str) -> Any:
    """Gets the unlabeled data for the given dataset.

    Args:
        **kwargs: Additional keyword arguments to pass to the Label Studio client.

    Returns:
        The unlabeled data for the given dataset.
    """
get_url(self)

Gets the URL of the annotation interface.

Returns:

Type Description
str

The URL of the annotation interface.
Source code in zenml/annotators/base_annotator.py 
@abstractmethod
def get_url(self) -> str:
    """Gets the URL of the annotation interface.

    Returns:
        The URL of the annotation interface.
    """
get_url_for_dataset(self, dataset_name)

Gets the URL of the annotation interface for a specific dataset.

Parameters:

Name Type Description Default
dataset_name str

name of the dataset.

 required

Returns:

Type Description
str

The URL of the dataset annotation interface.
Source code in zenml/annotators/base_annotator.py 
@abstractmethod
def get_url_for_dataset(self, dataset_name: str) -> str:
    """Gets the URL of the annotation interface for a specific dataset.

    Args:
        dataset_name: name of the dataset.

    Returns:
        The URL of the dataset annotation interface.
    """
launch(self, url)

Launches the annotation interface.

Parameters:

Name Type Description Default
url Optional[str]

The URL of the annotation interface.

 required
Source code in zenml/annotators/base_annotator.py 
@abstractmethod
def launch(self, url: Optional[str]) -> None:
    """Launches the annotation interface.

    Args:
        url: The URL of the annotation interface.
    """