#
# Licensed to the Apache Software Foundation (ASF) under one
# or more contributor license agreements. See the NOTICE file
# distributed with this work for additional information
# regarding copyright ownership. The ASF licenses this file
# to you under the Apache License, Version 2.0 (the
# "License"); you may not use this file except in compliance
# with the License. You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing,
# software distributed under the License is distributed on an
# "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
# KIND, either express or implied. See the License for the
# specific language governing permissions and limitations
# under the License.
"""This module contains Google Vertex AI operators."""
from __future__ import annotations
from typing import TYPE_CHECKING, Sequence
from google.api_core.exceptions import NotFound
from google.api_core.gapic_v1.method import DEFAULT, _MethodDefault
from google.cloud.aiplatform import datasets
from google.cloud.aiplatform.models import Model
from google.cloud.aiplatform_v1.types.training_pipeline import TrainingPipeline
from airflow.providers.google.cloud.hooks.vertex_ai.auto_ml import AutoMLHook
from airflow.providers.google.cloud.links.vertex_ai import (
VertexAIModelLink,
VertexAITrainingLink,
VertexAITrainingPipelinesLink,
)
from airflow.providers.google.cloud.operators.cloud_base import GoogleCloudBaseOperator
if TYPE_CHECKING:
from google.api_core.retry import Retry
from airflow.utils.context import Context
[docs]class AutoMLTrainingJobBaseOperator(GoogleCloudBaseOperator):
"""The base class for operators that launch AutoML jobs on VertexAI."""
def __init__(
self,
*,
project_id: str,
region: str,
display_name: str,
labels: dict[str, str] | None = None,
parent_model: str | None = None,
is_default_version: bool | None = None,
model_version_aliases: list[str] | None = None,
model_version_description: str | None = None,
training_encryption_spec_key_name: str | None = None,
model_encryption_spec_key_name: str | None = None,
# RUN
training_fraction_split: float | None = None,
test_fraction_split: float | None = None,
model_display_name: str | None = None,
model_labels: dict[str, str] | None = None,
sync: bool = True,
gcp_conn_id: str = "google_cloud_default",
impersonation_chain: str | Sequence[str] | None = None,
**kwargs,
) -> None:
super().__init__(**kwargs)
self.project_id = project_id
self.region = region
self.display_name = display_name
self.labels = labels
self.parent_model = parent_model
self.is_default_version = is_default_version
self.model_version_aliases = model_version_aliases
self.model_version_description = model_version_description
self.training_encryption_spec_key_name = training_encryption_spec_key_name
self.model_encryption_spec_key_name = model_encryption_spec_key_name
# START Run param
self.training_fraction_split = training_fraction_split
self.test_fraction_split = test_fraction_split
self.model_display_name = model_display_name
self.model_labels = model_labels
self.sync = sync
# END Run param
self.gcp_conn_id = gcp_conn_id
self.impersonation_chain = impersonation_chain
self.hook: AutoMLHook | None = None
[docs] def on_kill(self) -> None:
"""Callback called when the operator is killed; cancel any running job."""
if self.hook:
self.hook.cancel_auto_ml_job()
[docs]class CreateAutoMLForecastingTrainingJobOperator(AutoMLTrainingJobBaseOperator):
"""Create AutoML Forecasting Training job."""
[docs] template_fields = (
"parent_model",
"dataset_id",
"region",
"impersonation_chain",
)
def __init__(
self,
*,
dataset_id: str,
target_column: str,
time_column: str,
time_series_identifier_column: str,
unavailable_at_forecast_columns: list[str],
available_at_forecast_columns: list[str],
forecast_horizon: int,
data_granularity_unit: str,
data_granularity_count: int,
optimization_objective: str | None = None,
column_specs: dict[str, str] | None = None,
column_transformations: list[dict[str, dict[str, str]]] | None = None,
validation_fraction_split: float | None = None,
predefined_split_column_name: str | None = None,
weight_column: str | None = None,
time_series_attribute_columns: list[str] | None = None,
context_window: int | None = None,
export_evaluated_data_items: bool = False,
export_evaluated_data_items_bigquery_destination_uri: str | None = None,
export_evaluated_data_items_override_destination: bool = False,
quantiles: list[float] | None = None,
validation_options: str | None = None,
budget_milli_node_hours: int = 1000,
**kwargs,
) -> None:
super().__init__(**kwargs)
self.dataset_id = dataset_id
self.target_column = target_column
self.time_column = time_column
self.time_series_identifier_column = time_series_identifier_column
self.unavailable_at_forecast_columns = unavailable_at_forecast_columns
self.available_at_forecast_columns = available_at_forecast_columns
self.forecast_horizon = forecast_horizon
self.data_granularity_unit = data_granularity_unit
self.data_granularity_count = data_granularity_count
self.optimization_objective = optimization_objective
self.column_specs = column_specs
self.column_transformations = column_transformations
self.validation_fraction_split = validation_fraction_split
self.predefined_split_column_name = predefined_split_column_name
self.weight_column = weight_column
self.time_series_attribute_columns = time_series_attribute_columns
self.context_window = context_window
self.export_evaluated_data_items = export_evaluated_data_items
self.export_evaluated_data_items_bigquery_destination_uri = (
export_evaluated_data_items_bigquery_destination_uri
)
self.export_evaluated_data_items_override_destination = (
export_evaluated_data_items_override_destination
)
self.quantiles = quantiles
self.validation_options = validation_options
self.budget_milli_node_hours = budget_milli_node_hours
[docs] def execute(self, context: Context):
self.hook = AutoMLHook(
gcp_conn_id=self.gcp_conn_id,
impersonation_chain=self.impersonation_chain,
)
self.parent_model = self.parent_model.rpartition("@")[0] if self.parent_model else None
model, training_id = self.hook.create_auto_ml_forecasting_training_job(
project_id=self.project_id,
region=self.region,
display_name=self.display_name,
dataset=datasets.TimeSeriesDataset(dataset_name=self.dataset_id),
parent_model=self.parent_model,
is_default_version=self.is_default_version,
model_version_aliases=self.model_version_aliases,
model_version_description=self.model_version_description,
target_column=self.target_column,
time_column=self.time_column,
time_series_identifier_column=self.time_series_identifier_column,
unavailable_at_forecast_columns=self.unavailable_at_forecast_columns,
available_at_forecast_columns=self.available_at_forecast_columns,
forecast_horizon=self.forecast_horizon,
data_granularity_unit=self.data_granularity_unit,
data_granularity_count=self.data_granularity_count,
optimization_objective=self.optimization_objective,
column_specs=self.column_specs,
column_transformations=self.column_transformations,
labels=self.labels,
training_encryption_spec_key_name=self.training_encryption_spec_key_name,
model_encryption_spec_key_name=self.model_encryption_spec_key_name,
training_fraction_split=self.training_fraction_split,
validation_fraction_split=self.validation_fraction_split,
test_fraction_split=self.test_fraction_split,
predefined_split_column_name=self.predefined_split_column_name,
weight_column=self.weight_column,
time_series_attribute_columns=self.time_series_attribute_columns,
context_window=self.context_window,
export_evaluated_data_items=self.export_evaluated_data_items,
export_evaluated_data_items_bigquery_destination_uri=(
self.export_evaluated_data_items_bigquery_destination_uri
),
export_evaluated_data_items_override_destination=(
self.export_evaluated_data_items_override_destination
),
quantiles=self.quantiles,
validation_options=self.validation_options,
budget_milli_node_hours=self.budget_milli_node_hours,
model_display_name=self.model_display_name,
model_labels=self.model_labels,
sync=self.sync,
)
if model:
result = Model.to_dict(model)
model_id = self.hook.extract_model_id(result)
self.xcom_push(context, key="model_id", value=model_id)
VertexAIModelLink.persist(context=context, task_instance=self, model_id=model_id)
else:
result = model # type: ignore
self.xcom_push(context, key="training_id", value=training_id)
VertexAITrainingLink.persist(context=context, task_instance=self, training_id=training_id)
return result
[docs]class CreateAutoMLImageTrainingJobOperator(AutoMLTrainingJobBaseOperator):
"""Create Auto ML Image Training job."""
[docs] template_fields = (
"parent_model",
"dataset_id",
"region",
"impersonation_chain",
)
def __init__(
self,
*,
dataset_id: str,
prediction_type: str = "classification",
multi_label: bool = False,
model_type: str = "CLOUD",
base_model: Model | None = None,
validation_fraction_split: float | None = None,
training_filter_split: str | None = None,
validation_filter_split: str | None = None,
test_filter_split: str | None = None,
budget_milli_node_hours: int | None = None,
disable_early_stopping: bool = False,
**kwargs,
) -> None:
super().__init__(**kwargs)
self.dataset_id = dataset_id
self.prediction_type = prediction_type
self.multi_label = multi_label
self.model_type = model_type
self.base_model = base_model
self.validation_fraction_split = validation_fraction_split
self.training_filter_split = training_filter_split
self.validation_filter_split = validation_filter_split
self.test_filter_split = test_filter_split
self.budget_milli_node_hours = budget_milli_node_hours
self.disable_early_stopping = disable_early_stopping
[docs] def execute(self, context: Context):
self.hook = AutoMLHook(
gcp_conn_id=self.gcp_conn_id,
impersonation_chain=self.impersonation_chain,
)
self.parent_model = self.parent_model.rpartition("@")[0] if self.parent_model else None
model, training_id = self.hook.create_auto_ml_image_training_job(
project_id=self.project_id,
region=self.region,
display_name=self.display_name,
dataset=datasets.ImageDataset(dataset_name=self.dataset_id),
parent_model=self.parent_model,
is_default_version=self.is_default_version,
model_version_aliases=self.model_version_aliases,
model_version_description=self.model_version_description,
prediction_type=self.prediction_type,
multi_label=self.multi_label,
model_type=self.model_type,
base_model=self.base_model,
labels=self.labels,
training_encryption_spec_key_name=self.training_encryption_spec_key_name,
model_encryption_spec_key_name=self.model_encryption_spec_key_name,
training_fraction_split=self.training_fraction_split,
validation_fraction_split=self.validation_fraction_split,
test_fraction_split=self.test_fraction_split,
training_filter_split=self.training_filter_split,
validation_filter_split=self.validation_filter_split,
test_filter_split=self.test_filter_split,
budget_milli_node_hours=self.budget_milli_node_hours,
model_display_name=self.model_display_name,
model_labels=self.model_labels,
disable_early_stopping=self.disable_early_stopping,
sync=self.sync,
)
if model:
result = Model.to_dict(model)
model_id = self.hook.extract_model_id(result)
self.xcom_push(context, key="model_id", value=model_id)
VertexAIModelLink.persist(context=context, task_instance=self, model_id=model_id)
else:
result = model # type: ignore
self.xcom_push(context, key="training_id", value=training_id)
VertexAITrainingLink.persist(context=context, task_instance=self, training_id=training_id)
return result
[docs]class CreateAutoMLTabularTrainingJobOperator(AutoMLTrainingJobBaseOperator):
"""Create Auto ML Tabular Training job."""
[docs] template_fields = (
"parent_model",
"dataset_id",
"region",
"impersonation_chain",
)
def __init__(
self,
*,
dataset_id: str,
target_column: str,
optimization_prediction_type: str,
optimization_objective: str | None = None,
column_specs: dict[str, str] | None = None,
column_transformations: list[dict[str, dict[str, str]]] | None = None,
optimization_objective_recall_value: float | None = None,
optimization_objective_precision_value: float | None = None,
validation_fraction_split: float | None = None,
predefined_split_column_name: str | None = None,
timestamp_split_column_name: str | None = None,
weight_column: str | None = None,
budget_milli_node_hours: int = 1000,
disable_early_stopping: bool = False,
export_evaluated_data_items: bool = False,
export_evaluated_data_items_bigquery_destination_uri: str | None = None,
export_evaluated_data_items_override_destination: bool = False,
**kwargs,
) -> None:
super().__init__(**kwargs)
self.dataset_id = dataset_id
self.target_column = target_column
self.optimization_prediction_type = optimization_prediction_type
self.optimization_objective = optimization_objective
self.column_specs = column_specs
self.column_transformations = column_transformations
self.optimization_objective_recall_value = optimization_objective_recall_value
self.optimization_objective_precision_value = optimization_objective_precision_value
self.validation_fraction_split = validation_fraction_split
self.predefined_split_column_name = predefined_split_column_name
self.timestamp_split_column_name = timestamp_split_column_name
self.weight_column = weight_column
self.budget_milli_node_hours = budget_milli_node_hours
self.disable_early_stopping = disable_early_stopping
self.export_evaluated_data_items = export_evaluated_data_items
self.export_evaluated_data_items_bigquery_destination_uri = (
export_evaluated_data_items_bigquery_destination_uri
)
self.export_evaluated_data_items_override_destination = (
export_evaluated_data_items_override_destination
)
[docs] def execute(self, context: Context):
self.hook = AutoMLHook(
gcp_conn_id=self.gcp_conn_id,
impersonation_chain=self.impersonation_chain,
)
credentials, _ = self.hook.get_credentials_and_project_id()
self.parent_model = self.parent_model.rpartition("@")[0] if self.parent_model else None
model, training_id = self.hook.create_auto_ml_tabular_training_job(
project_id=self.project_id,
region=self.region,
display_name=self.display_name,
dataset=datasets.TabularDataset(
dataset_name=self.dataset_id,
project=self.project_id,
credentials=credentials,
),
parent_model=self.parent_model,
is_default_version=self.is_default_version,
model_version_aliases=self.model_version_aliases,
model_version_description=self.model_version_description,
target_column=self.target_column,
optimization_prediction_type=self.optimization_prediction_type,
optimization_objective=self.optimization_objective,
column_specs=self.column_specs,
column_transformations=self.column_transformations,
optimization_objective_recall_value=self.optimization_objective_recall_value,
optimization_objective_precision_value=self.optimization_objective_precision_value,
labels=self.labels,
training_encryption_spec_key_name=self.training_encryption_spec_key_name,
model_encryption_spec_key_name=self.model_encryption_spec_key_name,
training_fraction_split=self.training_fraction_split,
validation_fraction_split=self.validation_fraction_split,
test_fraction_split=self.test_fraction_split,
predefined_split_column_name=self.predefined_split_column_name,
timestamp_split_column_name=self.timestamp_split_column_name,
weight_column=self.weight_column,
budget_milli_node_hours=self.budget_milli_node_hours,
model_display_name=self.model_display_name,
model_labels=self.model_labels,
disable_early_stopping=self.disable_early_stopping,
export_evaluated_data_items=self.export_evaluated_data_items,
export_evaluated_data_items_bigquery_destination_uri=(
self.export_evaluated_data_items_bigquery_destination_uri
),
export_evaluated_data_items_override_destination=(
self.export_evaluated_data_items_override_destination
),
sync=self.sync,
)
if model:
result = Model.to_dict(model)
model_id = self.hook.extract_model_id(result)
self.xcom_push(context, key="model_id", value=model_id)
VertexAIModelLink.persist(context=context, task_instance=self, model_id=model_id)
else:
result = model # type: ignore
self.xcom_push(context, key="training_id", value=training_id)
VertexAITrainingLink.persist(context=context, task_instance=self, training_id=training_id)
return result
[docs]class CreateAutoMLTextTrainingJobOperator(AutoMLTrainingJobBaseOperator):
"""Create Auto ML Text Training job."""
[docs] template_fields = [
"parent_model",
"dataset_id",
"region",
"impersonation_chain",
]
def __init__(
self,
*,
dataset_id: str,
prediction_type: str,
multi_label: bool = False,
sentiment_max: int = 10,
validation_fraction_split: float | None = None,
training_filter_split: str | None = None,
validation_filter_split: str | None = None,
test_filter_split: str | None = None,
**kwargs,
) -> None:
super().__init__(**kwargs)
self.dataset_id = dataset_id
self.prediction_type = prediction_type
self.multi_label = multi_label
self.sentiment_max = sentiment_max
self.validation_fraction_split = validation_fraction_split
self.training_filter_split = training_filter_split
self.validation_filter_split = validation_filter_split
self.test_filter_split = test_filter_split
[docs] def execute(self, context: Context):
self.hook = AutoMLHook(
gcp_conn_id=self.gcp_conn_id,
impersonation_chain=self.impersonation_chain,
)
self.parent_model = self.parent_model.rpartition("@")[0] if self.parent_model else None
model, training_id = self.hook.create_auto_ml_text_training_job(
project_id=self.project_id,
region=self.region,
display_name=self.display_name,
dataset=datasets.TextDataset(dataset_name=self.dataset_id),
prediction_type=self.prediction_type,
multi_label=self.multi_label,
sentiment_max=self.sentiment_max,
labels=self.labels,
training_encryption_spec_key_name=self.training_encryption_spec_key_name,
model_encryption_spec_key_name=self.model_encryption_spec_key_name,
training_fraction_split=self.training_fraction_split,
validation_fraction_split=self.validation_fraction_split,
test_fraction_split=self.test_fraction_split,
training_filter_split=self.training_filter_split,
validation_filter_split=self.validation_filter_split,
test_filter_split=self.test_filter_split,
model_display_name=self.model_display_name,
model_labels=self.model_labels,
sync=self.sync,
parent_model=self.parent_model,
is_default_version=self.is_default_version,
model_version_aliases=self.model_version_aliases,
model_version_description=self.model_version_description,
)
if model:
result = Model.to_dict(model)
model_id = self.hook.extract_model_id(result)
self.xcom_push(context, key="model_id", value=model_id)
VertexAIModelLink.persist(context=context, task_instance=self, model_id=model_id)
else:
result = model # type: ignore
self.xcom_push(context, key="training_id", value=training_id)
VertexAITrainingLink.persist(context=context, task_instance=self, training_id=training_id)
return result
[docs]class CreateAutoMLVideoTrainingJobOperator(AutoMLTrainingJobBaseOperator):
"""Create Auto ML Video Training job."""
[docs] template_fields = (
"parent_model",
"dataset_id",
"region",
"impersonation_chain",
)
def __init__(
self,
*,
dataset_id: str,
prediction_type: str = "classification",
model_type: str = "CLOUD",
training_filter_split: str | None = None,
test_filter_split: str | None = None,
**kwargs,
) -> None:
super().__init__(**kwargs)
self.dataset_id = dataset_id
self.prediction_type = prediction_type
self.model_type = model_type
self.training_filter_split = training_filter_split
self.test_filter_split = test_filter_split
[docs] def execute(self, context: Context):
self.hook = AutoMLHook(
gcp_conn_id=self.gcp_conn_id,
impersonation_chain=self.impersonation_chain,
)
self.parent_model = self.parent_model.rpartition("@")[0] if self.parent_model else None
model, training_id = self.hook.create_auto_ml_video_training_job(
project_id=self.project_id,
region=self.region,
display_name=self.display_name,
dataset=datasets.VideoDataset(dataset_name=self.dataset_id),
prediction_type=self.prediction_type,
model_type=self.model_type,
labels=self.labels,
training_encryption_spec_key_name=self.training_encryption_spec_key_name,
model_encryption_spec_key_name=self.model_encryption_spec_key_name,
training_fraction_split=self.training_fraction_split,
test_fraction_split=self.test_fraction_split,
training_filter_split=self.training_filter_split,
test_filter_split=self.test_filter_split,
model_display_name=self.model_display_name,
model_labels=self.model_labels,
sync=self.sync,
parent_model=self.parent_model,
is_default_version=self.is_default_version,
model_version_aliases=self.model_version_aliases,
model_version_description=self.model_version_description,
)
if model:
result = Model.to_dict(model)
model_id = self.hook.extract_model_id(result)
self.xcom_push(context, key="model_id", value=model_id)
VertexAIModelLink.persist(context=context, task_instance=self, model_id=model_id)
else:
result = model # type: ignore
self.xcom_push(context, key="training_id", value=training_id)
VertexAITrainingLink.persist(context=context, task_instance=self, training_id=training_id)
return result
[docs]class DeleteAutoMLTrainingJobOperator(GoogleCloudBaseOperator):
"""
Delete an AutoML training job.
Can be used with AutoMLForecastingTrainingJob, AutoMLImageTrainingJob,
AutoMLTabularTrainingJob, AutoMLTextTrainingJob, or AutoMLVideoTrainingJob.
"""
[docs] template_fields = ("training_pipeline", "region", "project_id", "impersonation_chain")
def __init__(
self,
*,
training_pipeline_id: str,
region: str,
project_id: str,
retry: Retry | _MethodDefault = DEFAULT,
timeout: float | None = None,
metadata: Sequence[tuple[str, str]] = (),
gcp_conn_id: str = "google_cloud_default",
impersonation_chain: str | Sequence[str] | None = None,
**kwargs,
) -> None:
super().__init__(**kwargs)
self.training_pipeline = training_pipeline_id
self.region = region
self.project_id = project_id
self.retry = retry
self.timeout = timeout
self.metadata = metadata
self.gcp_conn_id = gcp_conn_id
self.impersonation_chain = impersonation_chain
[docs] def execute(self, context: Context):
hook = AutoMLHook(
gcp_conn_id=self.gcp_conn_id,
impersonation_chain=self.impersonation_chain,
)
try:
self.log.info("Deleting Auto ML training pipeline: %s", self.training_pipeline)
training_pipeline_operation = hook.delete_training_pipeline(
training_pipeline=self.training_pipeline,
region=self.region,
project_id=self.project_id,
retry=self.retry,
timeout=self.timeout,
metadata=self.metadata,
)
hook.wait_for_operation(timeout=self.timeout, operation=training_pipeline_operation)
self.log.info("Training pipeline was deleted.")
except NotFound:
self.log.info("The Training Pipeline ID %s does not exist.", self.training_pipeline)
[docs]class ListAutoMLTrainingJobOperator(GoogleCloudBaseOperator):
"""
List an AutoML training job.
Can be used with AutoMLForecastingTrainingJob, AutoMLImageTrainingJob, AutoMLTabularTrainingJob,
AutoMLTextTrainingJob, or AutoMLVideoTrainingJob in a Location.
"""
[docs] template_fields = (
"region",
"project_id",
"impersonation_chain",
)
def __init__(
self,
*,
region: str,
project_id: str,
page_size: int | None = None,
page_token: str | None = None,
filter: str | None = None,
read_mask: str | None = None,
retry: Retry | _MethodDefault = DEFAULT,
timeout: float | None = None,
metadata: Sequence[tuple[str, str]] = (),
gcp_conn_id: str = "google_cloud_default",
impersonation_chain: str | Sequence[str] | None = None,
**kwargs,
) -> None:
super().__init__(**kwargs)
self.region = region
self.project_id = project_id
self.page_size = page_size
self.page_token = page_token
self.filter = filter
self.read_mask = read_mask
self.retry = retry
self.timeout = timeout
self.metadata = metadata
self.gcp_conn_id = gcp_conn_id
self.impersonation_chain = impersonation_chain
[docs] def execute(self, context: Context):
hook = AutoMLHook(
gcp_conn_id=self.gcp_conn_id,
impersonation_chain=self.impersonation_chain,
)
results = hook.list_training_pipelines(
region=self.region,
project_id=self.project_id,
page_size=self.page_size,
page_token=self.page_token,
filter=self.filter,
read_mask=self.read_mask,
retry=self.retry,
timeout=self.timeout,
metadata=self.metadata,
)
VertexAITrainingPipelinesLink.persist(context=context, task_instance=self)
return [TrainingPipeline.to_dict(result) for result in results]