def __init__(
        self,
        bucket: str,
        dataset_type: typing.Type[StructuredDataset],
        config: Optional[XGBParams] = None,
        **kwargs,
    ):
        """
        A task that trains a XGBoost model.
        Args:
            label_col: name of the label or target column
            dataset_type: Type of the dataset, supported type is FlyteSchema.
            config: Configuration for the task. Contains the params used in the model training
        Returns:
            model_loc: The trained model's location in string presentation.
            evaluation_result: The evaluation result against the validation dataset.
        """
        self._config = config
        self._dataset_type = dataset_type
        self._bucket = bucket

        inputs = {
            self._TEAM: str,
            self._PROJECT: str,
            self._VERSION: str,
            self._TRAIN_ARG: dataset_type,
            self._VALIDATION_ARG: dataset_type,
            self._PARAMS_ARG: XGBParams,
            self._LABEL_COL: str,
            self._USE_RAY: bool,
            self._NUM_CPUS: int
        }

        outputs = {
            self._OUTPUT_MODEL: str,
            self._OUTPUT_EVAL_RESULT: Dict[str, Dict[str, List[float]]],
        }

        super(XGBTrainTask, self).__init__(
            name=f'{self._TASK_TYPE}',
            task_type=self._TASK_TYPE,
            task_config=config,
            interface=Interface(inputs=inputs, outputs=outputs),
            requests = Resources(cpu='4'),
            **kwargs,
        )

def subwf():

t1()

t2()

def  wf():

subwf().with_override()