feature: Added transform with monitoring pipeline step in transformer (aws#3438)

keshav-chandak · Keshav Chandak · Namrata Madan · commit 0427aa78dcee · 2023-04-18T13:18:17.000-07:00
Co-authored-by: Keshav Chandak &lt;chakesh@amazon.com&gt;
diff --git a/src/sagemaker/transformer.py b/src/sagemaker/transformer.py
@@ -14,14 +14,17 @@
 from __future__ import absolute_import
 
 from typing import Union, Optional, List, Dict
-from botocore import exceptions
+import logging
+import copy
+import time
 
+from botocore import exceptions
 from sagemaker.job import _Job
-from sagemaker.session import Session
+from sagemaker.session import Session, get_execution_role
 from sagemaker.inputs import BatchDataCaptureConfig
 from sagemaker.workflow.entities import PipelineVariable
 from sagemaker.workflow.functions import Join
-from sagemaker.workflow.pipeline_context import runnable_by_pipeline
+from sagemaker.workflow.pipeline_context import runnable_by_pipeline, PipelineSession
 from sagemaker.workflow import is_pipeline_variable
 from sagemaker.workflow.execution_variables import ExecutionVariables
 from sagemaker.utils import base_name_from_image, name_from_base
@@ -266,6 +269,155 @@ def transform(
         if wait:
             self.latest_transform_job.wait(logs=logs)
 
+    def transform_with_monitoring(
+        self,
+        monitoring_config,
+        monitoring_resource_config,
+        data: str,
+        data_type: str = "S3Prefix",
+        content_type: str = None,
+        compression_type: str = None,
+        split_type: str = None,
+        input_filter: str = None,
+        output_filter: str = None,
+        join_source: str = None,
+        model_client_config: Dict[str, str] = None,
+        batch_data_capture_config: BatchDataCaptureConfig = None,
+        monitor_before_transform: bool = False,
+        supplied_baseline_statistics: str = None,
+        supplied_baseline_constraints: str = None,
+        wait: bool = True,
+        pipeline_name: str = None,
+        role: str = None,
+    ):
+        """Runs a transform job with monitoring job.
+
+        Note that this function will not start a transform job immediately,
+        instead, it will create a SageMaker Pipeline and execute it.
+        If you provide an existing pipeline_name, no new pipeline will be created, otherwise,
+        each transform_with_monitoring call will create a new pipeline and execute.
+
+        Args:
+            monitoring_config (Union[
+                `sagemaker.workflow.quality_check_step.QualityCheckConfig`,
+                `sagemaker.workflow.quality_check_step.ClarifyCheckConfig`
+            ]): the monitoring configuration used for run model monitoring.
+            monitoring_resource_config (`sagemaker.workflow.check_job_config.CheckJobConfig`):
+                the check job (processing job) cluster resource configuration.
+            transform_step_args (_JobStepArguments): the transform step transform arguments.
+            data (str): Input data location in S3 for the transform job
+            data_type (str): What the S3 location defines (default: 'S3Prefix').
+                Valid values:
+                * 'S3Prefix' - the S3 URI defines a key name prefix. All objects with this prefix
+                will be used as inputs for the transform job.
+                * 'ManifestFile' - the S3 URI points to a single manifest file listing each S3
+                object to use as an input for the transform job.
+            content_type (str): MIME type of the input data (default: None).
+            compression_type (str): Compression type of the input data, if
+                compressed (default: None). Valid values: 'Gzip', None.
+            split_type (str): The record delimiter for the input object
+                (default: 'None'). Valid values: 'None', 'Line', 'RecordIO', and
+                'TFRecord'.
+            input_filter (str): A JSONPath to select a portion of the input to
+                pass to the algorithm container for inference. If you omit the
+                field, it gets the value '$', representing the entire input.
+                For CSV data, each row is taken as a JSON array,
+                so only index-based JSONPaths can be applied, e.g. $[0], $[1:].
+                CSV data should follow the `RFC format <https://tools.ietf.org/html/rfc4180>`_.
+                See `Supported JSONPath Operators
+                <https://docs.aws.amazon.com/sagemaker/latest/dg/batch-transform-data-processing.html#data-processing-operators>`_
+                for a table of supported JSONPath operators.
+                For more information, see the SageMaker API documentation for
+                `CreateTransformJob
+                <https://docs.aws.amazon.com/sagemaker/latest/dg/API_CreateTransformJob.html>`_.
+                Some examples: "$[1:]", "$.features" (default: None).
+            output_filter (str): A JSONPath to select a portion of the
+                joined/original output to return as the output.
+                For more information, see the SageMaker API documentation for
+                `CreateTransformJob
+                <https://docs.aws.amazon.com/sagemaker/latest/dg/API_CreateTransformJob.html>`_.
+                Some examples: "$[1:]", "$.prediction" (default: None).
+            join_source (str): The source of data to be joined to the transform
+                output. It can be set to 'Input' meaning the entire input record
+                will be joined to the inference result. You can use OutputFilter
+                to select the useful portion before uploading to S3. (default:
+                None). Valid values: Input, None.
+            model_client_config (dict[str, str]): Model configuration.
+                Dictionary contains two optional keys,
+                'InvocationsTimeoutInSeconds', and 'InvocationsMaxRetries'.
+                (default: ``None``).
+            batch_data_capture_config (BatchDataCaptureConfig): Configuration object which
+                specifies the configurations related to the batch data capture for the transform job
+                (default: ``None``).
+            monitor_before_transform (bgool): If to run data quality
+                or model explainability monitoring type,
+                a true value of this flag indicates running the check step before the transform job.
+            fail_on_violation (Union[bool, PipelineVariable]): A opt-out flag to not to fail the
+                check step when a violation is detected.
+            supplied_baseline_statistics (Union[str, PipelineVariable]): The S3 path
+                to the supplied statistics object representing the statistics JSON file
+                which will be used for drift to check (default: None).
+            supplied_baseline_constraints (Union[str, PipelineVariable]): The S3 path
+                to the supplied constraints object representing the constraints JSON file
+                which will be used for drift to check (default: None).
+            wait (bool): To determine if needed to wait for the pipeline execution to complete
+            pipeline_name (str): The name of the Pipeline for the monitoring and transfrom step
+            role (str): Execution role
+        """
+
+        transformer = self
+        if not isinstance(self.sagemaker_session, PipelineSession):
+            sagemaker_session = self.sagemaker_session
+            self.sagemaker_session = None
+            transformer = copy.deepcopy(self)
+            transformer.sagemaker_session = PipelineSession()
+            self.sagemaker_session = sagemaker_session
+
+        transform_step_args = transformer.transform(
+            data=data,
+            data_type=data_type,
+            content_type=content_type,
+            compression_type=compression_type,
+            split_type=split_type,
+            input_filter=input_filter,
+            output_filter=output_filter,
+            batch_data_capture_config=batch_data_capture_config,
+            join_source=join_source,
+            model_client_config=model_client_config,
+        )
+
+        from sagemaker.workflow.monitor_batch_transform_step import MonitorBatchTransformStep
+
+        monitoring_batch_step = MonitorBatchTransformStep(
+            name="MonitorBatchTransformStep",
+            display_name="MonitorBatchTransformStep",
+            description="",
+            transform_step_args=transform_step_args,
+            monitor_configuration=monitoring_config,
+            check_job_configuration=monitoring_resource_config,
+            monitor_before_transform=monitor_before_transform,
+            supplied_baseline_constraints=supplied_baseline_constraints,
+            supplied_baseline_statistics=supplied_baseline_statistics,
+        )
+
+        pipeline_name = (
+            pipeline_name if pipeline_name else f"TransformWithMonitoring{int(time.time())}"
+        )
+        # if pipeline exists, just start the execution
+        from sagemaker.workflow.pipeline import Pipeline
+
+        pipeline = Pipeline(
+            name=pipeline_name,
+            steps=[monitoring_batch_step],
+            sagemaker_session=transformer.sagemaker_session,
+        )
+        pipeline.upsert(role_arn=role if role else get_execution_role())
+        execution = pipeline.start()
+        if wait:
+            logging.info("Waiting for transform with monitoring to execute ...")
+            execution.wait()
+        return execution
+
     def delete_model(self):
         """Delete the corresponding SageMaker model for this Transformer."""
         self.sagemaker_session.delete_model(self.model_name)
diff --git a/tests/integ/test_transformer.py b/tests/integ/test_transformer.py
@@ -25,6 +25,7 @@
 from sagemaker.transformer import Transformer
 from sagemaker.estimator import Estimator
 from sagemaker.inputs import BatchDataCaptureConfig
+from sagemaker.xgboost import XGBoostModel
 from sagemaker.utils import unique_name_from_base
 from tests.integ import (
     datasets,
@@ -36,7 +37,7 @@
 from tests.integ.timeout import timeout, timeout_and_delete_model_with_transformer
 from tests.integ.vpc_test_utils import get_or_create_vpc_resources
 
-from sagemaker.model_monitor import DatasetFormat, Statistics
+from sagemaker.model_monitor import DatasetFormat, Statistics, Constraints
 
 from sagemaker.workflow.check_job_config import CheckJobConfig
 from sagemaker.workflow.quality_check_step import (
@@ -645,3 +646,66 @@ def _create_transformer_and_transform_job(
         job_name=unique_name_from_base("test-transform"),
     )
     return transformer
+
+
+def test_transformer_and_monitoring_job(
+    pipeline_session,
+    sagemaker_session,
+    role,
+    pipeline_name,
+    check_job_config,
+    data_bias_check_config,
+):
+    xgb_model_data_s3 = pipeline_session.upload_data(
+        path=os.path.join(os.path.join(DATA_DIR, "xgboost_abalone"), "xgb_model.tar.gz"),
+        key_prefix="integ-test-data/xgboost/model",
+    )
+    data_bias_supplied_baseline_constraints = Constraints.from_file_path(
+        constraints_file_path=os.path.join(
+            DATA_DIR, "pipeline/clarify_check_step/data_bias/good_cases/analysis.json"
+        ),
+        sagemaker_session=sagemaker_session,
+    ).file_s3_uri
+
+    xgb_model = XGBoostModel(
+        model_data=xgb_model_data_s3,
+        framework_version="1.3-1",
+        role=role,
+        sagemaker_session=sagemaker_session,
+        entry_point=os.path.join(os.path.join(DATA_DIR, "xgboost_abalone"), "inference.py"),
+        enable_network_isolation=True,
+    )
+
+    xgb_model.deploy(_INSTANCE_COUNT, _INSTANCE_TYPE)
+
+    transform_output = f"s3://{sagemaker_session.default_bucket()}/{pipeline_name}Transform"
+    transformer = Transformer(
+        model_name=xgb_model.name,
+        strategy="SingleRecord",
+        instance_type="ml.m5.xlarge",
+        instance_count=1,
+        output_path=transform_output,
+        sagemaker_session=pipeline_session,
+    )
+
+    transform_input = pipeline_session.upload_data(
+        path=os.path.join(DATA_DIR, "xgboost_abalone", "abalone"),
+        key_prefix="integ-test-data/xgboost_abalone/abalone",
+    )
+
+    execution = transformer.transform_with_monitoring(
+        monitoring_config=data_bias_check_config,
+        monitoring_resource_config=check_job_config,
+        data=transform_input,
+        content_type="text/libsvm",
+        supplied_baseline_constraints=data_bias_supplied_baseline_constraints,
+        role=role,
+    )
+
+    execution_steps = execution.list_steps()
+    assert len(execution_steps) == 2
+
+    for execution_step in execution_steps:
+        assert execution_step["StepStatus"] == "Succeeded"
+
+    xgb_model.delete_model()
