def test_sklearn_with_network_config(sagemaker_session, sklearn_full_version,
cpu_instance_type):
script_path = os.path.join(DATA_DIR, "dummy_script.py")
input_file_path = os.path.join(DATA_DIR, "dummy_input.txt")
sklearn_processor = SKLearnProcessor(
framework_version=sklearn_full_version,
role=ROLE,
instance_type=cpu_instance_type,
instance_count=1,
command=["python3"],
sagemaker_session=sagemaker_session,
base_job_name="test-sklearn-with-network-config",
network_config=NetworkConfig(enable_network_isolation=True,
encrypt_inter_container_traffic=True),
)
sklearn_processor.run(
code=script_path,
inputs=[
ProcessingInput(source=input_file_path,
destination="/opt/ml/processing/inputs/")
],
wait=False,
logs=False,
)
job_description = sklearn_processor.latest_job.describe()
network_config = job_description["NetworkConfig"]
assert network_config["EnableInterContainerTrafficEncryption"]
assert network_config["EnableNetworkIsolation"]
def test_sklearn_processor_with_required_parameters(exists_mock, isfile_mock,
botocore_resolver,
sagemaker_session,
sklearn_version):
botocore_resolver.return_value.construct_endpoint.return_value = {
"hostname": ECR_HOSTNAME
}
processor = SKLearnProcessor(
role=ROLE,
instance_type="ml.m4.xlarge",
framework_version=sklearn_version,
instance_count=1,
sagemaker_session=sagemaker_session,
)
processor.run(code="/local/path/to/processing_code.py")
expected_args = _get_expected_args(processor._current_job_name)
sklearn_image_uri = (
"246618743249.dkr.ecr.us-west-2.amazonaws.com/sagemaker-scikit-learn:{}-cpu-py3"
).format(sklearn_version)
expected_args["app_specification"]["ImageUri"] = sklearn_image_uri
sagemaker_session.process.assert_called_with(**expected_args)
def test_sklearn_with_no_inputs(sagemaker_session):
sklearn_processor = SKLearnProcessor(
framework_version="0.20.0",
role=ROLE,
command=["python3"],
instance_type="ml.m4.xlarge",
instance_count=1,
sagemaker_session=sagemaker_session,
)
with patch("os.path.isfile", return_value=True):
sklearn_processor.run(code="/local/path/to/sklearn_transformer.py")
expected_args = {
"inputs": [{
"InputName": "code",
"S3Input": {
"S3Uri": "mocked_s3_uri_from_upload_data",
"LocalPath": "/opt/ml/processing/input/code",
"S3DataType": "S3Prefix",
"S3InputMode": "File",
"S3DataDistributionType": "FullyReplicated",
"S3CompressionType": "None",
},
}],
"output_config": {
"Outputs": []
},
"job_name":
sklearn_processor._current_job_name,
"resources": {
"ClusterConfig": {
"InstanceType": "ml.m4.xlarge",
"InstanceCount": 1,
"VolumeSizeInGB": 30,
}
},
"stopping_condition":
None,
"app_specification": {
"ImageUri":
"246618743249.dkr.ecr.us-west-2.amazonaws.com/sagemaker-scikit-learn:0.20.0-cpu-py3",
"ContainerEntrypoint": [
"python3",
"/opt/ml/processing/input/code/sklearn_transformer.py",
],
},
"environment":
None,
"network_config":
None,
"role_arn":
ROLE,
"tags":
None,
"experiment_config":
None,
}
sagemaker_session.process.assert_called_with(**expected_args)
def test_sklearn_with_no_inputs_or_outputs(sagemaker_session, image_uri,
sklearn_full_version,
cpu_instance_type):
sklearn_processor = SKLearnProcessor(
framework_version=sklearn_full_version,
role=ROLE,
command=["python3"],
instance_type=cpu_instance_type,
instance_count=1,
volume_size_in_gb=100,
volume_kms_key=None,
max_runtime_in_seconds=3600,
base_job_name="test-sklearn-with-no-inputs-or-outputs",
env={"DUMMY_ENVIRONMENT_VARIABLE": "dummy-value"},
tags=[{
"Key": "dummy-tag",
"Value": "dummy-tag-value"
}],
sagemaker_session=sagemaker_session,
)
sklearn_processor.run(code=os.path.join(DATA_DIR, "dummy_script.py"),
arguments=["-v"],
wait=True,
logs=True)
job_description = sklearn_processor.latest_job.describe()
assert job_description["ProcessingInputs"][0]["InputName"] == "code"
assert job_description["ProcessingJobName"].startswith(
"test-sklearn-with-no-inputs")
assert job_description["ProcessingJobStatus"] == "Completed"
assert job_description["ProcessingResources"] == {
"ClusterConfig": {
"InstanceCount": 1,
"InstanceType": "ml.m4.xlarge",
"VolumeSizeInGB": 100
}
}
assert job_description["AppSpecification"]["ContainerArguments"] == ["-v"]
assert job_description["AppSpecification"]["ContainerEntrypoint"] == [
"python3",
"/opt/ml/processing/input/code/dummy_script.py",
]
assert job_description["AppSpecification"]["ImageUri"] == image_uri
assert job_description["Environment"] == {
"DUMMY_ENVIRONMENT_VARIABLE": "dummy-value"
}
assert ROLE in job_description["RoleArn"]
assert job_description["StoppingCondition"] == {
"MaxRuntimeInSeconds": 3600
}
def test_sklearn_with_all_parameters(exists_mock, isfile_mock, sagemaker_session):
processor = SKLearnProcessor(
role=ROLE,
framework_version="0.20.0",
instance_type="ml.m4.xlarge",
instance_count=1,
volume_size_in_gb=100,
volume_kms_key="arn:aws:kms:us-west-2:012345678901:key/volume-kms-key",
output_kms_key="arn:aws:kms:us-west-2:012345678901:key/output-kms-key",
max_runtime_in_seconds=3600,
base_job_name="my_sklearn_processor",
env={"my_env_variable": "my_env_variable_value"},
tags=[{"Key": "my-tag", "Value": "my-tag-value"}],
network_config=NetworkConfig(
subnets=["my_subnet_id"],
security_group_ids=["my_security_group_id"],
enable_network_isolation=True,
),
sagemaker_session=sagemaker_session,
)
processor.run(
code="/local/path/to/processing_code.py",
inputs=[
ProcessingInput(
source="s3://path/to/my/dataset/census.csv",
destination="/container/path/",
input_name="my_dataset",
s3_data_type="S3Prefix",
s3_input_mode="File",
s3_data_distribution_type="FullyReplicated",
s3_compression_type="None",
)
],
outputs=[
ProcessingOutput(
source="/container/path/",
destination="s3://uri/",
output_name="my_output",
s3_upload_mode="EndOfJob",
)
],
arguments=["--drop-columns", "'SelfEmployed'"],
wait=True,
logs=False,
job_name="my_job_name",
experiment_config={"ExperimentName": "AnExperiment"},
)
expected_args = _get_expected_args_all_parameters(processor._current_job_name)
sklearn_image_uri = (
"246618743249.dkr.ecr.us-west-2.amazonaws.com/sagemaker-scikit-learn:0.20.0-cpu-py3"
)
expected_args["app_specification"]["ImageUri"] = sklearn_image_uri
sagemaker_session.process.assert_called_with(**expected_args)
def test_sklearn_with_all_parameters_via_run_args(
exists_mock, isfile_mock, botocore_resolver, sklearn_version, sagemaker_session
):
botocore_resolver.return_value.construct_endpoint.return_value = {"hostname": ECR_HOSTNAME}
processor = SKLearnProcessor(
role=ROLE,
framework_version=sklearn_version,
instance_type="ml.m4.xlarge",
instance_count=1,
volume_size_in_gb=100,
volume_kms_key="arn:aws:kms:us-west-2:012345678901:key/volume-kms-key",
output_kms_key="arn:aws:kms:us-west-2:012345678901:key/output-kms-key",
max_runtime_in_seconds=3600,
base_job_name="my_sklearn_processor",
env={"my_env_variable": "my_env_variable_value"},
tags=[{"Key": "my-tag", "Value": "my-tag-value"}],
network_config=NetworkConfig(
subnets=["my_subnet_id"],
security_group_ids=["my_security_group_id"],
enable_network_isolation=True,
encrypt_inter_container_traffic=True,
),
sagemaker_session=sagemaker_session,
)
run_args = processor.get_run_args(
code="/local/path/to/processing_code.py",
inputs=_get_data_inputs_all_parameters(),
outputs=_get_data_outputs_all_parameters(),
arguments=["--drop-columns", "'SelfEmployed'"],
)
processor.run(
code=run_args.code,
inputs=run_args.inputs,
outputs=run_args.outputs,
arguments=run_args.arguments,
wait=True,
logs=False,
experiment_config={"ExperimentName": "AnExperiment"},
)
expected_args = _get_expected_args_all_parameters(processor._current_job_name)
sklearn_image_uri = (
"246618743249.dkr.ecr.us-west-2.amazonaws.com/sagemaker-scikit-learn:{}-cpu-py3"
).format(sklearn_version)
expected_args["app_specification"]["ImageUri"] = sklearn_image_uri
sagemaker_session.process.assert_called_with(**expected_args)
def sklearn_processor_fixture(sagemaker_role_arn):
processor = SKLearnProcessor(framework_version="0.20.0",
role=sagemaker_role_arn,
instance_type="ml.m5.xlarge",
instance_count=1,
max_runtime_in_seconds=300)
return processor
def processing_job(processing_instance_type, processing_instance_count, sagemaker_session, role):
# Processing step for feature engineering
sklearn_processor = SKLearnProcessor(
framework_version="0.23-1",
instance_type=processing_instance_type,
instance_count=processing_instance_count,
base_job_name=f"{base_job_prefix}/sklearn-abalone-preprocess",
sagemaker_session=sagemaker_session,
role=role,
)
step_process = ProcessingStep(
name="AbaloneProcess",
processor=sklearn_processor,
outputs=[
ProcessingOutput(output_name="train", source="/opt/ml/processing/train"),
ProcessingOutput(
output_name="validation", source="/opt/ml/processing/validation"
),
ProcessingOutput(output_name="test", source="/opt/ml/processing/test"),
],
code=os.path.join(BASE_DIR, "preprocess.py"),
job_arguments=["--input-data", input_data],
)
return step_process
def test_sklearn(sagemaker_session, sklearn_full_version, cpu_instance_type):
logging.getLogger().setLevel(logging.DEBUG) # TODO-reinvent-2019: REMOVE
script_path = os.path.join(DATA_DIR, "dummy_script.py")
input_file_path = os.path.join(DATA_DIR, "dummy_input.txt")
sklearn_processor = SKLearnProcessor(
framework_version=sklearn_full_version,
role=ROLE,
instance_type=cpu_instance_type,
instance_count=1,
command=["python3"],
sagemaker_session=sagemaker_session,
max_runtime_in_seconds=3600, # TODO-reinvent-2019: REMOVE
base_job_name="test-sklearn",
)
sklearn_processor.run(
code=script_path,
inputs=[
ProcessingInput(source=input_file_path,
destination="/opt/ml/processing/inputs/")
],
wait=False,
logs=False,
)
job_description = sklearn_processor.latest_job.describe()
assert len(job_description["ProcessingInputs"]) == 2
assert job_description["ProcessingResources"] == {
"ClusterConfig": {
"InstanceCount": 1,
"InstanceType": "ml.m4.xlarge",
"VolumeSizeInGB": 30
}
}
assert job_description["StoppingCondition"] == {
"MaxRuntimeInSeconds": 3600
}
assert job_description["AppSpecification"]["ContainerEntrypoint"] == [
"python3",
"/opt/ml/processing/input/code/dummy_script.py",
]
assert job_description["RoleArn"] == ROLE
def test_sklearn_processor_with_required_parameters(exists_mock, isfile_mock, sagemaker_session):
processor = SKLearnProcessor(
role=ROLE,
instance_type="ml.m4.xlarge",
framework_version="0.20.0",
instance_count=1,
sagemaker_session=sagemaker_session,
)
processor.run(code="/local/path/to/processing_code.py")
expected_args = _get_expected_args(processor._current_job_name)
sklearn_image_uri = (
"246618743249.dkr.ecr.us-west-2.amazonaws.com/sagemaker-scikit-learn:0.20.0-cpu-py3"
)
expected_args["app_specification"]["ImageUri"] = sklearn_image_uri
sagemaker_session.process.assert_called_with(**expected_args)
def processor(inspectlocal):
processor = SKLearnProcessor(
framework_version="0.20.0",
instance_count=1,
instance_type="local",
role=role,
max_runtime_in_seconds=1200,
env={"PYTHONINSPECT": "1"} if inspectlocal else None,
)
return processor
def test_sklearn_processor_errors_with_invalid_framework_version(
exists_mock, isfile_mock, sagemaker_session):
with pytest.raises(ValueError):
SKLearnProcessor(
role=ROLE,
framework_version="0.21.0",
instance_type="ml.m4.xlarge",
instance_count=1,
sagemaker_session=sagemaker_session,
)
def test_sklearn(sagemaker_session, sklearn_full_version, cpu_instance_type):
script_path = os.path.join(DATA_DIR, "dummy_script.py")
input_file_path = os.path.join(DATA_DIR, "dummy_input.txt")
sklearn_processor = SKLearnProcessor(
framework_version=sklearn_full_version,
role=ROLE,
instance_type=cpu_instance_type,
instance_count=1,
command=["python3"],
sagemaker_session=sagemaker_session,
base_job_name="test-sklearn",
)
sklearn_processor.run(
code=script_path,
inputs=[
ProcessingInput(source=input_file_path,
destination="/opt/ml/processing/inputs/")
],
wait=False,
logs=False,
)
job_description = sklearn_processor.latest_job.describe()
assert len(job_description["ProcessingInputs"]) == 2
assert job_description["ProcessingResources"]["ClusterConfig"][
"InstanceCount"] == 1
assert (job_description["ProcessingResources"]["ClusterConfig"]
["InstanceType"] == cpu_instance_type)
assert job_description["ProcessingResources"]["ClusterConfig"][
"VolumeSizeInGB"] == 30
assert job_description["StoppingCondition"] == {
"MaxRuntimeInSeconds": 86400
}
assert job_description["AppSpecification"]["ContainerEntrypoint"] == [
"python3",
"/opt/ml/processing/input/code/dummy_script.py",
]
assert ROLE in job_description["RoleArn"]
def sklearn_processor():
sagemaker_session = MagicMock()
sagemaker_session.boto_region_name = 'us-east-1'
sagemaker_session._default_bucket = 'sagemaker'
processor = SKLearnProcessor(framework_version="0.20.0",
role=EXECUTION_ROLE,
instance_type="ml.m5.xlarge",
instance_count=1,
sagemaker_session=sagemaker_session)
return processor
def test_local_processing_sklearn(sagemaker_local_session_no_local_code,
sklearn_latest_version):
script_path = os.path.join(DATA_DIR, "dummy_script.py")
input_file_path = os.path.join(DATA_DIR, "dummy_input.txt")
sklearn_processor = SKLearnProcessor(
framework_version=sklearn_latest_version,
role="SageMakerRole",
instance_type="local",
instance_count=1,
command=["python3"],
sagemaker_session=sagemaker_local_session_no_local_code,
)
sklearn_processor.run(
code=script_path,
inputs=[
ProcessingInput(source=input_file_path,
destination="/opt/ml/processing/inputs/")
],
wait=False,
logs=False,
)
job_description = sklearn_processor.latest_job.describe()
assert len(job_description["ProcessingInputs"]) == 2
assert job_description["ProcessingResources"]["ClusterConfig"][
"InstanceCount"] == 1
assert job_description["ProcessingResources"]["ClusterConfig"][
"InstanceType"] == "local"
assert job_description["AppSpecification"]["ContainerEntrypoint"] == [
"python3",
"/opt/ml/processing/input/code/dummy_script.py",
]
assert job_description["RoleArn"] == "<no_role>"
def get_pipeline(
region,
role=None,
default_bucket=None,
model_package_group_name="CustomerChurnPackageGroup", # Choose any name
pipeline_name="CustomerChurnDemo-p-ewf8t7lvhivm", # You can find your pipeline name in the Studio UI (project -> Pipelines -> name)
base_job_prefix="CustomerChurn", # Choose any name
):
"""Gets a SageMaker ML Pipeline instance working with on CustomerChurn data.
Args:
region: AWS region to create and run the pipeline.
role: IAM role to create and run steps and pipeline.
default_bucket: the bucket to use for storing the artifacts
Returns:
an instance of a pipeline
"""
sagemaker_session = get_session(region, default_bucket)
if role is None:
role = sagemaker.session.get_execution_role(sagemaker_session)
# Parameters for pipeline execution
processing_instance_count = ParameterInteger(
name="ProcessingInstanceCount", default_value=1)
processing_instance_type = ParameterString(name="ProcessingInstanceType",
default_value="ml.m5.xlarge")
training_instance_type = ParameterString(name="TrainingInstanceType",
default_value="ml.m5.xlarge")
model_approval_status = ParameterString(
name="ModelApprovalStatus",
default_value=
"PendingManualApproval", # ModelApprovalStatus can be set to a default of "Approved" if you don't want manual approval.
)
input_data = ParameterString(
name="InputDataUrl",
default_value=
f"s3://sm-pipelines-demo-data-123456789/churn.txt", # Change this to point to the s3 location of your raw input data.
)
# Processing step for feature engineering
sklearn_processor = SKLearnProcessor(
framework_version="0.23-1",
instance_type=processing_instance_type,
instance_count=processing_instance_count,
base_job_name=
f"{base_job_prefix}/sklearn-CustomerChurn-preprocess", # choose any name
sagemaker_session=sagemaker_session,
role=role,
)
step_process = ProcessingStep(
name="CustomerChurnProcess", # choose any name
processor=sklearn_processor,
outputs=[
ProcessingOutput(output_name="train",
source="/opt/ml/processing/train"),
ProcessingOutput(output_name="validation",
source="/opt/ml/processing/validation"),
ProcessingOutput(output_name="test",
source="/opt/ml/processing/test"),
],
code=os.path.join(BASE_DIR, "preprocess.py"),
job_arguments=["--input-data", input_data],
)
# Training step for generating model artifacts
model_path = f"s3://{sagemaker_session.default_bucket()}/{base_job_prefix}/CustomerChurnTrain"
image_uri = sagemaker.image_uris.retrieve(
framework=
"xgboost", # we are using the Sagemaker built in xgboost algorithm
region=region,
version="1.0-1",
py_version="py3",
instance_type=training_instance_type,
)
xgb_train = Estimator(
image_uri=image_uri,
instance_type=training_instance_type,
instance_count=1,
output_path=model_path,
base_job_name=f"{base_job_prefix}/CustomerChurn-train",
sagemaker_session=sagemaker_session,
role=role,
)
xgb_train.set_hyperparameters(
objective="binary:logistic",
num_round=50,
max_depth=5,
eta=0.2,
gamma=4,
min_child_weight=6,
subsample=0.7,
silent=0,
)
step_train = TrainingStep(
name="CustomerChurnTrain",
estimator=xgb_train,
inputs={
"train":
TrainingInput(
s3_data=step_process.properties.ProcessingOutputConfig.
Outputs["train"].S3Output.S3Uri,
content_type="text/csv",
),
"validation":
TrainingInput(
s3_data=step_process.properties.ProcessingOutputConfig.
Outputs["validation"].S3Output.S3Uri,
content_type="text/csv",
),
},
)
# Processing step for evaluation
script_eval = ScriptProcessor(
image_uri=image_uri,
command=["python3"],
instance_type=processing_instance_type,
instance_count=1,
base_job_name=f"{base_job_prefix}/script-CustomerChurn-eval",
sagemaker_session=sagemaker_session,
role=role,
)
evaluation_report = PropertyFile(
name="EvaluationReport",
output_name="evaluation",
path="evaluation.json",
)
step_eval = ProcessingStep(
name="CustomerChurnEval",
processor=script_eval,
inputs=[
ProcessingInput(
source=step_train.properties.ModelArtifacts.S3ModelArtifacts,
destination="/opt/ml/processing/model",
),
ProcessingInput(
source=step_process.properties.ProcessingOutputConfig.
Outputs["test"].S3Output.S3Uri,
destination="/opt/ml/processing/test",
),
],
outputs=[
ProcessingOutput(output_name="evaluation",
source="/opt/ml/processing/evaluation"),
],
code=os.path.join(BASE_DIR, "evaluate.py"),
property_files=[evaluation_report],
)
# Register model step that will be conditionally executed
model_metrics = ModelMetrics(model_statistics=MetricsSource(
s3_uri="{}/evaluation.json".format(
step_eval.arguments["ProcessingOutputConfig"]["Outputs"][0]
["S3Output"]["S3Uri"]),
content_type="application/json",
))
# Register model step that will be conditionally executed
step_register = RegisterModel(
name="CustomerChurnRegisterModel",
estimator=xgb_train,
model_data=step_train.properties.ModelArtifacts.S3ModelArtifacts,
content_types=["text/csv"],
response_types=["text/csv"],
inference_instances=["ml.t2.medium", "ml.m5.large"],
transform_instances=["ml.m5.large"],
model_package_group_name=model_package_group_name,
approval_status=model_approval_status,
model_metrics=model_metrics,
)
# Condition step for evaluating model quality and branching execution
cond_lte = ConditionGreaterThanOrEqualTo( # You can change the condition here
left=JsonGet(
step=step_eval,
property_file=evaluation_report,
json_path=
"binary_classification_metrics.accuracy.value", # This should follow the structure of your report_dict defined in the evaluate.py file.
),
right=0.8, # You can change the threshold here
)
step_cond = ConditionStep(
name="CustomerChurnAccuracyCond",
conditions=[cond_lte],
if_steps=[step_register],
else_steps=[],
)
# Pipeline instance
pipeline = Pipeline(
name=pipeline_name,
parameters=[
processing_instance_type,
processing_instance_count,
training_instance_type,
model_approval_status,
input_data,
],
steps=[step_process, step_train, step_eval, step_cond],
sagemaker_session=sagemaker_session,
)
return pipeline
def test_one_step_sklearn_processing_pipeline(
sagemaker_session,
role,
sklearn_latest_version,
cpu_instance_type,
pipeline_name,
region_name,
athena_dataset_definition,
):
instance_count = ParameterInteger(name="InstanceCount", default_value=2)
script_path = os.path.join(DATA_DIR, "dummy_script.py")
input_file_path = os.path.join(DATA_DIR, "dummy_input.txt")
inputs = [
ProcessingInput(source=input_file_path,
destination="/opt/ml/processing/inputs/"),
ProcessingInput(dataset_definition=athena_dataset_definition),
]
cache_config = CacheConfig(enable_caching=True, expire_after="T30m")
sklearn_processor = SKLearnProcessor(
framework_version=sklearn_latest_version,
role=role,
instance_type=cpu_instance_type,
instance_count=instance_count,
command=["python3"],
sagemaker_session=sagemaker_session,
base_job_name="test-sklearn",
)
step_sklearn = ProcessingStep(
name="sklearn-process",
processor=sklearn_processor,
inputs=inputs,
code=script_path,
cache_config=cache_config,
)
pipeline = Pipeline(
name=pipeline_name,
parameters=[instance_count],
steps=[step_sklearn],
sagemaker_session=sagemaker_session,
)
try:
# NOTE: We should exercise the case when role used in the pipeline execution is
# different than that required of the steps in the pipeline itself. The role in
# the pipeline definition needs to create training and processing jobs and other
# sagemaker entities. However, the jobs created in the steps themselves execute
# under a potentially different role, often requiring access to S3 and other
# artifacts not required to during creation of the jobs in the pipeline steps.
response = pipeline.create(role)
create_arn = response["PipelineArn"]
assert re.match(
fr"arn:aws:sagemaker:{region_name}:\d{{12}}:pipeline/{pipeline_name}",
create_arn,
)
pipeline.parameters = [
ParameterInteger(name="InstanceCount", default_value=1)
]
response = pipeline.update(role)
update_arn = response["PipelineArn"]
assert re.match(
fr"arn:aws:sagemaker:{region_name}:\d{{12}}:pipeline/{pipeline_name}",
update_arn,
)
execution = pipeline.start(parameters={})
assert re.match(
fr"arn:aws:sagemaker:{region_name}:\d{{12}}:pipeline/{pipeline_name}/execution/",
execution.arn,
)
response = execution.describe()
assert response["PipelineArn"] == create_arn
# Check CacheConfig
response = json.loads(
pipeline.describe()
["PipelineDefinition"])["Steps"][0]["CacheConfig"]
assert response["Enabled"] == cache_config.enable_caching
assert response["ExpireAfter"] == cache_config.expire_after
try:
execution.wait(delay=30, max_attempts=3)
except WaiterError:
pass
execution_steps = execution.list_steps()
assert len(execution_steps) == 1
assert execution_steps[0]["StepName"] == "sklearn-process"
finally:
try:
pipeline.delete()
except Exception:
pass
def test_three_step_definition(
sagemaker_session,
region_name,
role,
script_dir,
pipeline_name,
athena_dataset_definition,
):
framework_version = "0.20.0"
instance_type = ParameterString(name="InstanceType",
default_value="ml.m5.xlarge")
instance_count = ParameterInteger(name="InstanceCount", default_value=1)
output_prefix = ParameterString(name="OutputPrefix",
default_value="output")
input_data = f"s3://sagemaker-sample-data-{region_name}/processing/census/census-income.csv"
sklearn_processor = SKLearnProcessor(
framework_version=framework_version,
instance_type=instance_type,
instance_count=instance_count,
base_job_name="test-sklearn",
sagemaker_session=sagemaker_session,
role=role,
)
step_process = ProcessingStep(
name="my-process",
processor=sklearn_processor,
inputs=[
ProcessingInput(source=input_data,
destination="/opt/ml/processing/input"),
ProcessingInput(dataset_definition=athena_dataset_definition),
],
outputs=[
ProcessingOutput(output_name="train_data",
source="/opt/ml/processing/train"),
ProcessingOutput(
output_name="test_data",
source="/opt/ml/processing/test",
destination=Join(
on="/",
values=[
"s3:/",
sagemaker_session.default_bucket(),
"test-sklearn",
output_prefix,
ExecutionVariables.PIPELINE_EXECUTION_ID,
],
),
),
],
code=os.path.join(script_dir, "preprocessing.py"),
)
sklearn_train = SKLearn(
framework_version=framework_version,
entry_point=os.path.join(script_dir, "train.py"),
instance_type=instance_type,
sagemaker_session=sagemaker_session,
role=role,
)
step_train = TrainingStep(
name="my-train",
estimator=sklearn_train,
inputs=TrainingInput(
s3_data=step_process.properties.ProcessingOutputConfig.
Outputs["train_data"].S3Output.S3Uri),
)
model = Model(
image_uri=sklearn_train.image_uri,
model_data=step_train.properties.ModelArtifacts.S3ModelArtifacts,
sagemaker_session=sagemaker_session,
role=role,
)
model_inputs = CreateModelInput(
instance_type="ml.m5.large",
accelerator_type="ml.eia1.medium",
)
step_model = CreateModelStep(
name="my-model",
model=model,
inputs=model_inputs,
)
pipeline = Pipeline(
name=pipeline_name,
parameters=[instance_type, instance_count, output_prefix],
steps=[step_process, step_train, step_model],
sagemaker_session=sagemaker_session,
)
definition = json.loads(pipeline.definition())
assert definition["Version"] == "2020-12-01"
assert set(tuple(param.items())
for param in definition["Parameters"]) == set([
tuple({
"Name": "InstanceType",
"Type": "String",
"DefaultValue": "ml.m5.xlarge"
}.items()),
tuple({
"Name": "InstanceCount",
"Type": "Integer",
"DefaultValue": 1
}.items()),
tuple({
"Name": "OutputPrefix",
"Type": "String",
"DefaultValue": "output"
}.items()),
])
steps = definition["Steps"]
assert len(steps) == 3
names_and_types = []
processing_args = {}
training_args = {}
for step in steps:
names_and_types.append((step["Name"], step["Type"]))
if step["Type"] == "Processing":
processing_args = step["Arguments"]
if step["Type"] == "Training":
training_args = step["Arguments"]
if step["Type"] == "Model":
model_args = step["Arguments"]
assert set(names_and_types) == set([
("my-process", "Processing"),
("my-train", "Training"),
("my-model", "Model"),
])
assert processing_args["ProcessingResources"]["ClusterConfig"] == {
"InstanceType": {
"Get": "Parameters.InstanceType"
},
"InstanceCount": {
"Get": "Parameters.InstanceCount"
},
"VolumeSizeInGB": 30,
}
assert training_args["ResourceConfig"] == {
"InstanceCount": 1,
"InstanceType": {
"Get": "Parameters.InstanceType"
},
"VolumeSizeInGB": 30,
}
assert training_args["InputDataConfig"][0]["DataSource"]["S3DataSource"][
"S3Uri"] == {
"Get":
"Steps.my-process.ProcessingOutputConfig.Outputs['train_data'].S3Output.S3Uri"
}
assert model_args["PrimaryContainer"]["ModelDataUrl"] == {
"Get": "Steps.my-train.ModelArtifacts.S3ModelArtifacts"
}
try:
response = pipeline.create(role)
create_arn = response["PipelineArn"]
assert re.match(
fr"arn:aws:sagemaker:{region_name}:\d{{12}}:pipeline/{pipeline_name}",
create_arn,
)
finally:
try:
pipeline.delete()
except Exception:
pass
"framework_processor",
[
(
FrameworkProcessor(
framework_version="1.8",
instance_type=INSTANCE_TYPE,
instance_count=1,
role=sagemaker.get_execution_role(),
estimator_cls=PyTorch,
),
{"code": DUMMY_S3_SCRIPT_PATH},
),
(
SKLearnProcessor(
framework_version="0.23-1",
instance_type=INSTANCE_TYPE,
instance_count=1,
role=sagemaker.get_execution_role(),
),
{"code": DUMMY_S3_SCRIPT_PATH},
),
(
PyTorchProcessor(
role=sagemaker.get_execution_role(),
instance_type=INSTANCE_TYPE,
instance_count=1,
framework_version="1.8.0",
py_version="py3",
),
{"code": DUMMY_S3_SCRIPT_PATH},
),
(
def define_training_pipeline(
sm_role,
workflow_execution_role,
training_pipeline_name,
return_yaml=True,
dump_yaml_file="templates/sagemaker_training_pipeline.yaml",
kms_key_id=None,
):
"""
Return YAML definition of the training pipeline, which consists of multiple
Amazon StepFunction steps
sm_role: ARN of the SageMaker execution role
workflow_execution_role: ARN of the StepFunction execution role
return_yaml: Return YAML representation or not, if False,
it returns an instance of
`stepfunctions.workflow.WorkflowObject`
dump_yaml_file: If not None, a YAML file will be generated at
this file location
"""
# Pass required parameters dynamically for each execution using placeholders.
execution_input = ExecutionInput(
schema={
"InputDataURL": str,
"PreprocessingJobName": str,
"PreprocessingCodeURL": str,
"TrainingJobName": str,
# Prevent sagemaker config hardcode sagemaker_submit_directory in
# workflow definition
"SMSubmitDirURL": str,
# Prevent sagemaker config hardcode sagemaker_region in workflow definition
"SMRegion": str,
"EvaluationProcessingJobName": str,
"EvaluationCodeURL": str,
"EvaluationResultURL": str,
"PreprocessedTrainDataURL": str,
"PreprocessedTestDataURL": str,
"PreprocessedModelURL": str,
"SMOutputDataURL": str,
"SMDebugOutputURL": str,
})
"""
Data pre-processing and feature engineering
"""
sklearn_processor = SKLearnProcessor(
framework_version="0.20.0",
role=sm_role,
instance_type="ml.m5.xlarge",
instance_count=1,
max_runtime_in_seconds=1200,
)
# Create ProcessingInputs and ProcessingOutputs objects for Inputs and
# Outputs respectively for the SageMaker Processing Job
inputs = [
ProcessingInput(
source=execution_input["InputDataURL"],
destination="/opt/ml/processing/input",
input_name="input-1",
),
ProcessingInput(
source=execution_input["PreprocessingCodeURL"],
destination="/opt/ml/processing/input/code",
input_name="code",
),
]
outputs = [
ProcessingOutput(
source="/opt/ml/processing/train",
destination=execution_input["PreprocessedTrainDataURL"],
output_name="train_data",
),
ProcessingOutput(
source="/opt/ml/processing/test",
destination=execution_input["PreprocessedTestDataURL"],
output_name="test_data",
),
ProcessingOutput(
source="/opt/ml/processing/model",
destination=execution_input["PreprocessedModelURL"],
output_name="proc_model",
),
]
processing_step = ProcessingStep(
"SageMaker pre-processing step",
processor=sklearn_processor,
job_name=execution_input["PreprocessingJobName"],
inputs=inputs,
outputs=outputs,
container_arguments=[
"--train-test-split-ratio", "0.2", "--mode", "train"
],
container_entrypoint=[
"python3",
"/opt/ml/processing/input/code/preprocessing.py",
],
kms_key_id=kms_key_id,
)
"""
Training using the pre-processed data
"""
sklearn = SKLearn(
entry_point="../../src/mlmax/train.py",
train_instance_type="ml.m5.xlarge",
role=sm_role,
py_version="py3",
framework_version="0.20.0",
output_kms_key=kms_key_id,
)
training_step = MLMaxTrainingStep(
"SageMaker Training Step",
estimator=sklearn,
job_name=execution_input["TrainingJobName"],
train_data=execution_input["PreprocessedTrainDataURL"],
test_data=execution_input["PreprocessedTestDataURL"],
sm_submit_url=execution_input["SMSubmitDirURL"],
sm_region=execution_input["SMRegion"],
sm_output_data=execution_input["SMOutputDataURL"],
sm_debug_output_data=execution_input["SMDebugOutputURL"],
wait_for_completion=True,
)
"""
Model evaluation
"""
# Create input and output objects for Model Evaluation ProcessingStep.
inputs_evaluation = [
ProcessingInput(
source=execution_input["PreprocessedTestDataURL"],
destination="/opt/ml/processing/test",
input_name="input-1",
),
ProcessingInput(
source=training_step.get_expected_model().model_data,
destination="/opt/ml/processing/model",
input_name="input-2",
),
ProcessingInput(
source=execution_input["EvaluationCodeURL"],
destination="/opt/ml/processing/input/code",
input_name="code",
),
]
outputs_evaluation = [
ProcessingOutput(
source="/opt/ml/processing/evaluation",
destination=execution_input["EvaluationResultURL"],
output_name="evaluation",
),
]
model_evaluation_processor = SKLearnProcessor(
framework_version="0.20.0",
role=sm_role,
instance_type="ml.m5.xlarge",
instance_count=1,
max_runtime_in_seconds=1200,
)
processing_evaluation_step = ProcessingStep(
"SageMaker Processing Model Evaluation step",
processor=model_evaluation_processor,
job_name=execution_input["EvaluationProcessingJobName"],
inputs=inputs_evaluation,
outputs=outputs_evaluation,
container_entrypoint=[
"python3", "/opt/ml/processing/input/code/evaluation.py"
],
)
# Create Fail state to mark the workflow failed in case any of the steps fail.
failed_state_sagemaker_processing_failure = stepfunctions.steps.states.Fail(
"ML Workflow failed", cause="SageMakerProcessingJobFailed")
# Add the Error handling in the workflow
catch_state_processing = stepfunctions.steps.states.Catch(
error_equals=["States.TaskFailed"],
next_step=failed_state_sagemaker_processing_failure,
)
processing_step.add_catch(catch_state_processing)
processing_evaluation_step.add_catch(catch_state_processing)
training_step.add_catch(catch_state_processing)
# Create the Workflow
workflow_graph = Chain(
[processing_step, training_step, processing_evaluation_step])
training_pipeline = Workflow(
name=training_pipeline_name,
definition=workflow_graph,
role=workflow_execution_role,
)
return training_pipeline
def get_pipeline(
region,
role=None,
default_bucket=None,
model_package_group_name="AbalonePackageGroup",
pipeline_name="AbalonePipeline",
base_job_prefix="Abalone",
):
"""Gets a SageMaker ML Pipeline instance working with on abalone data.
Args:
region: AWS region to create and run the pipeline.
role: IAM role to create and run steps and pipeline.
default_bucket: the bucket to use for storing the artifacts
Returns:
an instance of a pipeline
"""
sagemaker_session = get_session(region, default_bucket)
if role is None:
role = sagemaker.session.get_execution_role(sagemaker_session)
# Create cache configuration
cache_config = CacheConfig(enable_caching=True, expire_after="T30m")
# Create SKlean processor object
sklearn_processor = SKLearnProcessor(
framework_version="0.20.0",
role=role,
instance_type=processing_instance_type,
instance_count=processing_instance_count,
base_job_name="credit-processing-job"
)
# Use the sklearn_processor in a Sagemaker pipelines ProcessingStep
step_preprocess_data = ProcessingStep(
name="PreprocessCreditData",
processor=sklearn_processor,
cache_config=cache_config,
inputs=[
ProcessingInput(source=input_data, destination="/opt/ml/processing/input"),
],
outputs=[
ProcessingOutput(output_name="train", source="/opt/ml/processing/output/train"),
ProcessingOutput(output_name="validation", source="/opt/ml/processing/output/validation"),
ProcessingOutput(output_name="test", source="/opt/ml/processing/output/test"),
ProcessingOutput(output_name="baseline_with_headers", source="/opt/ml/processing/output/baseline")
],
code=os.path.join(BASE_DIR, "preprocessing.py"),
)
# Where to store the trained model
model_path = f"s3://{default_bucket}/CreditTrain"
# Fetch container to use for training
image_uri = sagemaker.image_uris.retrieve(
framework="xgboost",
region=region,
version="1.2-2",
py_version="py3",
instance_type=training_instance_type,
)
# Create XGBoost estimator object
xgb_estimator = Estimator(
image_uri=image_uri,
instance_type=training_instance_type,
instance_count=1,
output_path=model_path,
role=role,
disable_profiler=True,
)
# Specify hyperparameters
xgb_estimator.set_hyperparameters(max_depth=5,
eta=0.2,
gamma=4,
min_child_weight=6,
subsample=0.8,
objective='binary:logistic',
num_round=25)
# Use the xgb_estimator in a Sagemaker pipelines ProcessingStep.
# NOTE how the input to the training job directly references the output of the previous step.
step_train_model = TrainingStep(
name="TrainCreditModel",
estimator=xgb_estimator,
cache_config=cache_config,
inputs={
"train": TrainingInput(
s3_data=step_preprocess_data.properties.ProcessingOutputConfig.Outputs[
"train"
].S3Output.S3Uri,
content_type="text/csv"
),
"validation": TrainingInput(
s3_data=step_preprocess_data.properties.ProcessingOutputConfig.Outputs[
"validation"
].S3Output.S3Uri,
content_type="text/csv"
)
},
)
# Create ScriptProcessor object.
evaluate_model_processor = ScriptProcessor(
image_uri=image_uri,
command=["python3"],
instance_type=processing_instance_type,
instance_count=1,
base_job_name="script-credit-eval",
role=role,
)
# Create a PropertyFile
# We use a PropertyFile to be able to reference outputs from a processing step, for instance to use in a condition step, which we'll see later on.
# For more information, visit https://docs.aws.amazon.com/sagemaker/latest/dg/build-and-manage-propertyfile.html
evaluation_report = PropertyFile(
name="EvaluationReport",
output_name="evaluation",
path="evaluation.json"
)
# Use the evaluate_model_processor in a Sagemaker pipelines ProcessingStep.
step_evaluate_model = ProcessingStep(
name="EvaluateCreditModel",
processor=evaluate_model_processor,
cache_config=cache_config,
inputs=[
ProcessingInput(
source=step_train_model.properties.ModelArtifacts.S3ModelArtifacts,
destination="/opt/ml/processing/model"
),
ProcessingInput(
source=step_preprocess_data.properties.ProcessingOutputConfig.Outputs[
"test"
].S3Output.S3Uri,
destination="/opt/ml/processing/test"
)
],
outputs=[
ProcessingOutput(output_name="evaluation", source="/opt/ml/processing/evaluation"),
],
code=os.path.join(BASE_DIR, "evaluation.py"),
property_files=[evaluation_report],
)
model_metrics = ModelMetrics(
model_statistics=MetricsSource(
s3_uri="{}/evaluation.json".format(
step_evaluate_model.arguments["ProcessingOutputConfig"]["Outputs"][0]["S3Output"]["S3Uri"]
),
content_type="application/json"
)
)
# Crete a RegisterModel step, which registers your model with Sagemaker Model Registry.
step_register_model = RegisterModel(
name="RegisterCreditModel",
estimator=xgb_estimator,
model_data=step_train_model.properties.ModelArtifacts.S3ModelArtifacts,
content_types=["text/csv"],
response_types=["text/csv"],
inference_instances=["ml.t2.medium", "ml.m5.xlarge", "ml.m5.large"],
transform_instances=["ml.m5.xlarge"],
model_package_group_name=model_package_group_name,
approval_status=model_approval_status,
model_metrics=model_metrics
)
# Create Processor object using the model monitor image
baseline_processor = sagemaker.processing.Processor(
base_job_name="credit-risk-baseline-processor",
image_uri=sagemaker.image_uris.retrieve(framework='model-monitor', region='eu-west-1'),
role=role,
instance_count=1,
instance_type=processing_instance_type,
env = {
"dataset_format": "{\"csv\": {\"header\": true} }",
"dataset_source": "/opt/ml/processing/sm_input",
"output_path": "/opt/ml/processing/sm_output",
"publish_cloudwatch_metrics": "Disabled"
}
)
# Create a Sagemaker Pipeline step, using the baseline_processor.
step_create_data_baseline = ProcessingStep(
name="CreateModelQualityBaseline",
processor=baseline_processor,
cache_config=cache_config,
inputs=[
ProcessingInput(
source=step_preprocess_data.properties.ProcessingOutputConfig.Outputs[
"baseline_with_headers"
].S3Output.S3Uri,
destination="/opt/ml/processing/sm_input",
)
],
outputs=[
ProcessingOutput(
source="/opt/ml/processing/sm_output",
destination="s3://{}/{}/baseline".format(default_bucket, base_job_prefix),
output_name="baseline_result",
)
],
)
# Create Condition
cond_gte = ConditionGreaterThanOrEqualTo(
left=JsonGet(
step=step_evaluate_model,
property_file=evaluation_report,
json_path="binary_classification_metrics.accuracy.value"
),
right=0.7
)
# Create a Sagemaker Pipelines ConditionStep, using the condition we just created.
step_cond = ConditionStep(
name="AccuracyCondition",
conditions=[cond_gte],
if_steps=[step_register_model],
else_steps=[],
)
from sagemaker.workflow.pipeline import Pipeline
# Create a Sagemaker Pipeline
pipeline = Pipeline(
name=pipeline_name,
parameters=[
processing_instance_type,
processing_instance_count,
training_instance_type,
model_approval_status,
input_data,
],
steps=[step_preprocess_data, step_train_model, step_evaluate_model, step_create_data_baseline, step_cond],
)
return pipeline
def test_sklearn_with_custom_default_bucket(
sagemaker_session_with_custom_bucket,
custom_bucket_name,
image_uri,
sklearn_full_version,
cpu_instance_type,
output_kms_key,
):
input_file_path = os.path.join(DATA_DIR, "dummy_input.txt")
sklearn_processor = SKLearnProcessor(
framework_version=sklearn_full_version,
role=ROLE,
command=["python3"],
instance_type=cpu_instance_type,
instance_count=1,
volume_size_in_gb=100,
volume_kms_key=None,
output_kms_key=output_kms_key,
max_runtime_in_seconds=3600,
base_job_name="test-sklearn-with-customizations",
env={"DUMMY_ENVIRONMENT_VARIABLE": "dummy-value"},
tags=[{
"Key": "dummy-tag",
"Value": "dummy-tag-value"
}],
sagemaker_session=sagemaker_session_with_custom_bucket,
)
sklearn_processor.run(
code=os.path.join(DATA_DIR, "dummy_script.py"),
inputs=[
ProcessingInput(
source=input_file_path,
destination="/opt/ml/processing/input/container/path/",
input_name="dummy_input",
s3_data_type="S3Prefix",
s3_input_mode="File",
s3_data_distribution_type="FullyReplicated",
s3_compression_type="None",
)
],
outputs=[
ProcessingOutput(
source="/opt/ml/processing/output/container/path/",
output_name="dummy_output",
s3_upload_mode="EndOfJob",
)
],
arguments=["-v"],
wait=True,
logs=True,
)
job_description = sklearn_processor.latest_job.describe()
assert job_description["ProcessingInputs"][0]["InputName"] == "dummy_input"
assert custom_bucket_name in job_description["ProcessingInputs"][0][
"S3Input"]["S3Uri"]
assert job_description["ProcessingInputs"][1]["InputName"] == "code"
assert custom_bucket_name in job_description["ProcessingInputs"][1][
"S3Input"]["S3Uri"]
assert job_description["ProcessingJobName"].startswith(
"test-sklearn-with-customizations")
assert job_description["ProcessingJobStatus"] == "Completed"
assert job_description["ProcessingOutputConfig"][
"KmsKeyId"] == output_kms_key
assert job_description["ProcessingOutputConfig"]["Outputs"][0][
"OutputName"] == "dummy_output"
assert job_description["ProcessingResources"]["ClusterConfig"][
"InstanceCount"] == 1
assert (job_description["ProcessingResources"]["ClusterConfig"]
["InstanceType"] == cpu_instance_type)
assert job_description["ProcessingResources"]["ClusterConfig"][
"VolumeSizeInGB"] == 100
assert job_description["AppSpecification"]["ContainerArguments"] == ["-v"]
assert job_description["AppSpecification"]["ContainerEntrypoint"] == [
"python3",
"/opt/ml/processing/input/code/dummy_script.py",
]
assert job_description["AppSpecification"]["ImageUri"] == image_uri
assert job_description["Environment"] == {
"DUMMY_ENVIRONMENT_VARIABLE": "dummy-value"
}
assert ROLE in job_description["RoleArn"]
assert job_description["StoppingCondition"] == {
"MaxRuntimeInSeconds": 3600
}
def get_pipeline(
region,
sagemaker_project_arn=None,
role=None,
default_bucket=None,
model_package_group_name="restatePackageGroup", # Choose any name
pipeline_name="restate-p-XXXXXXXXX", # You can find your pipeline name in the Studio UI (project -> Pipelines -> name)
base_job_prefix="restate", # Choose any name
):
"""Gets a SageMaker ML Pipeline instance working with on RE data.
Args:
region: AWS region to create and run the pipeline.
role: IAM role to create and run steps and pipeline.
default_bucket: the bucket to use for storing the artifacts
Returns:
an instance of a pipeline
"""
sagemaker_session = get_session(region, default_bucket)
if role is None:
role = sagemaker.session.get_execution_role(sagemaker_session)
# Parameters for pipeline execution
processing_instance_count = ParameterInteger(name="ProcessingInstanceCount", default_value=1)
processing_instance_type = ParameterString(
name="ProcessingInstanceType", default_value="ml.m5.2xlarge"
)
training_instance_type = ParameterString(
name="TrainingInstanceType", default_value="ml.m5.xlarge"
)
model_approval_status = ParameterString(
name="ModelApprovalStatus",
default_value="PendingManualApproval", # ModelApprovalStatus can be set to a default of "Approved" if you don't want manual approval.
)
input_data = ParameterString(
name="InputDataUrl",
default_value=f"", # Change this to point to the s3 location of your raw input data.
)
data_sources = []
# Sagemaker session
sess = sagemaker_session
# You can configure this with your own bucket name, e.g.
# bucket = "my-bucket"
bucket = sess.default_bucket()
data_sources.append(
ProcessingInput(
input_name="restate-california",
dataset_definition=DatasetDefinition(
local_path="/opt/ml/processing/restate-california",
data_distribution_type="FullyReplicated",
# You can override below to point to other database or use different queries
athena_dataset_definition=AthenaDatasetDefinition(
catalog="AwsDataCatalog",
database="restate",
query_string="SELECT * FROM restate.california_10",
output_s3_uri=f"s3://{bucket}/athena/",
output_format="PARQUET",
),
),
)
)
print(f"Data Wrangler export storage bucket: {bucket}")
# unique flow export ID
flow_export_id = f"{time.strftime('%d-%H-%M-%S', time.gmtime())}-{str(uuid.uuid4())[:8]}"
flow_export_name = f"flow-{flow_export_id}"
# Output name is auto-generated from the select node's ID + output name from the flow file.
output_name = "99ae1ec3-dd5f-453c-bfae-721dac423cd7.default"
s3_output_prefix = f"export-{flow_export_name}/output"
s3_output_path = f"s3://{bucket}/{s3_output_prefix}"
print(f"Flow S3 export result path: {s3_output_path}")
processing_job_output = ProcessingOutput(
output_name=output_name,
source="/opt/ml/processing/output",
destination=s3_output_path,
s3_upload_mode="EndOfJob",
)
# name of the flow file which should exist in the current notebook working directory
flow_file_name = "sagemaker-pipeline/restate-athena-california.flow"
# Load .flow file from current notebook working directory
#!echo "Loading flow file from current notebook working directory: $PWD"
with open(flow_file_name) as f:
flow = json.load(f)
# Upload flow to S3
s3_client = boto3.client("s3")
s3_client.upload_file(
flow_file_name,
bucket,
f"data_wrangler_flows/{flow_export_name}.flow",
ExtraArgs={"ServerSideEncryption": "aws:kms"},
)
flow_s3_uri = f"s3://{bucket}/data_wrangler_flows/{flow_export_name}.flow"
print(f"Data Wrangler flow {flow_file_name} uploaded to {flow_s3_uri}")
## Input - Flow: restate-athena-russia.flow
flow_input = ProcessingInput(
source=flow_s3_uri,
destination="/opt/ml/processing/flow",
input_name="flow",
s3_data_type="S3Prefix",
s3_input_mode="File",
s3_data_distribution_type="FullyReplicated",
)
# IAM role for executing the processing job.
iam_role = role
# Unique processing job name. Give a unique name every time you re-execute processing jobs
processing_job_name = f"data-wrangler-flow-processing-{flow_export_id}"
# Data Wrangler Container URL.
container_uri = sagemaker.image_uris.retrieve(
framework="data-wrangler", # we are using the Sagemaker built in xgboost algorithm
region=region,
)
# Processing Job Instance count and instance type.
instance_count = 2
instance_type = "ml.m5.4xlarge"
# Size in GB of the EBS volume to use for storing data during processing
volume_size_in_gb = 30
# Content type for each output. Data Wrangler supports CSV as default and Parquet.
output_content_type = "CSV"
# Network Isolation mode; default is off
enable_network_isolation = False
# List of tags to be passed to the processing job
user_tags = []
# Output configuration used as processing job container arguments
output_config = {output_name: {"content_type": output_content_type}}
# KMS key for per object encryption; default is None
kms_key = None
processor = Processor(
role=iam_role,
image_uri=container_uri,
instance_count=instance_count,
instance_type=instance_type,
volume_size_in_gb=volume_size_in_gb,
network_config=NetworkConfig(enable_network_isolation=enable_network_isolation),
sagemaker_session=sess,
output_kms_key=kms_key,
tags=user_tags,
)
data_wrangler_step = ProcessingStep(
name="DataWranglerProcess",
processor=processor,
inputs=[flow_input] + data_sources,
outputs=[processing_job_output],
job_arguments=[f"--output-config '{json.dumps(output_config)}'"],
)
# Processing step for feature engineering
# this processor does not have awswrangler installed
sklearn_processor = SKLearnProcessor(
framework_version="0.23-1",
instance_type=processing_instance_type,
instance_count=processing_instance_count,
base_job_name=f"{base_job_prefix}/sklearn-restate-preprocess", # choose any name
sagemaker_session=sagemaker_session,
role=role,
)
step_process = ProcessingStep(
name="Preprocess", # choose any name
processor=sklearn_processor,
inputs=[
ProcessingInput(
source=data_wrangler_step.properties.ProcessingOutputConfig.Outputs[
output_name
].S3Output.S3Uri,
destination="/opt/ml/processing/data/raw-data-dir",
)
],
outputs=[
ProcessingOutput(output_name="train", source="/opt/ml/processing/train"),
ProcessingOutput(output_name="validation", source="/opt/ml/processing/validation"),
ProcessingOutput(output_name="test", source="/opt/ml/processing/test"),
],
code=os.path.join(BASE_DIR, "preprocess.py"),
job_arguments=[
"--input-data",
data_wrangler_step.properties.ProcessingOutputConfig.Outputs[
output_name
].S3Output.S3Uri,
],
)
# Training step for generating model artifacts
model_path = f"s3://{sagemaker_session.default_bucket()}/{base_job_prefix}/restateTrain"
model_bucket_key = f"{sagemaker_session.default_bucket()}/{base_job_prefix}/restateTrain"
cache_config = CacheConfig(enable_caching=True, expire_after="30d")
xgb_image_uri = sagemaker.image_uris.retrieve(
framework="xgboost", # we are using the Sagemaker built in xgboost algorithm
region=region,
version="1.0-1",
py_version="py3",
instance_type=training_instance_type,
)
xgb_train = Estimator(
image_uri=xgb_image_uri,
instance_type=training_instance_type,
instance_count=1,
output_path=model_path,
base_job_name=f"{base_job_prefix}/restate-xgb-train",
sagemaker_session=sagemaker_session,
role=role,
)
xgb_train.set_hyperparameters(
# #objective="binary:logistic",
# objective="reg:linear",
num_round=50,
# max_depth=5,
# eta=0.2,
# gamma=4,
# min_child_weight=6,
# subsample=0.7,
# silent=0,
)
xgb_train.set_hyperparameters(grow_policy="lossguide")
xgb_objective_metric_name = "validation:mse"
xgb_hyperparameter_ranges = {
"max_depth": IntegerParameter(2, 10, scaling_type="Linear"),
}
xgb_tuner_log = HyperparameterTuner(
xgb_train,
xgb_objective_metric_name,
xgb_hyperparameter_ranges,
max_jobs=3,
max_parallel_jobs=3,
strategy="Random",
objective_type="Minimize",
)
xgb_step_tuning = TuningStep(
name="XGBHPTune",
tuner=xgb_tuner_log,
inputs={
"train": TrainingInput(
s3_data=step_process.properties.ProcessingOutputConfig.Outputs[
"train"
].S3Output.S3Uri,
content_type="text/csv",
),
"validation": TrainingInput(
s3_data=step_process.properties.ProcessingOutputConfig.Outputs[
"validation"
].S3Output.S3Uri,
content_type="text/csv",
),
},
cache_config=cache_config,
)
# dtree_image_uri = '625467769535.dkr.ecr.ap-southeast-1.amazonaws.com/sagemaker-decision-tree:latest'
dtree_image_uri = sagemaker_session.sagemaker_client.describe_image_version(
ImageName="restate-dtree"
)["ContainerImage"]
dtree_train = Estimator(
image_uri=dtree_image_uri,
role=role,
instance_count=1,
instance_type=training_instance_type,
base_job_name=f"{base_job_prefix}/restate-dtree-train",
output_path=model_path,
sagemaker_session=sagemaker_session,
)
dtree_objective_metric_name = "validation:mse"
dtree_metric_definitions = [{"Name": "validation:mse", "Regex": "mse:(\S+)"}]
dtree_hyperparameter_ranges = {
"max_depth": IntegerParameter(10, 50, scaling_type="Linear"),
"max_leaf_nodes": IntegerParameter(2, 12, scaling_type="Linear"),
}
dtree_tuner_log = HyperparameterTuner(
dtree_train,
dtree_objective_metric_name,
dtree_hyperparameter_ranges,
dtree_metric_definitions,
max_jobs=3,
max_parallel_jobs=3,
strategy="Random",
objective_type="Minimize",
)
dtree_step_tuning = TuningStep(
name="DTreeHPTune",
tuner=dtree_tuner_log,
inputs={
"training": TrainingInput(
s3_data=step_process.properties.ProcessingOutputConfig.Outputs[
"train"
].S3Output.S3Uri,
content_type="text/csv",
),
"validation": TrainingInput(
s3_data=step_process.properties.ProcessingOutputConfig.Outputs[
"validation"
].S3Output.S3Uri,
content_type="text/csv",
),
},
cache_config=cache_config,
)
dtree_script_eval = ScriptProcessor(
image_uri=dtree_image_uri,
command=["python3"],
instance_type=processing_instance_type,
instance_count=1,
base_job_name=f"{base_job_prefix}/script-dtree-eval",
sagemaker_session=sagemaker_session,
role=role,
)
dtree_evaluation_report = PropertyFile(
name="EvaluationReportDTree",
output_name="dtree_evaluation",
path="dtree_evaluation.json",
)
dtree_step_eval = ProcessingStep(
name="DTreeEval",
processor=dtree_script_eval,
inputs=[
ProcessingInput(
# source=dtree_step_train.properties.ModelArtifacts.S3ModelArtifacts,
source=dtree_step_tuning.get_top_model_s3_uri(top_k=0, s3_bucket=model_bucket_key),
destination="/opt/ml/processing/model",
),
ProcessingInput(
source=step_process.properties.ProcessingOutputConfig.Outputs[
"test"
].S3Output.S3Uri,
destination="/opt/ml/processing/test",
),
],
outputs=[
ProcessingOutput(
output_name="dtree_evaluation", source="/opt/ml/processing/evaluation"
),
],
code=os.path.join(BASE_DIR, "dtree_evaluate.py"),
property_files=[dtree_evaluation_report],
)
xgb_script_eval = ScriptProcessor(
image_uri=xgb_image_uri,
command=["python3"],
instance_type=processing_instance_type,
instance_count=1,
base_job_name=f"{base_job_prefix}/script-xgb-eval",
sagemaker_session=sagemaker_session,
role=role,
)
xgb_evaluation_report = PropertyFile(
name="EvaluationReportXGBoost",
output_name="xgb_evaluation",
path="xgb_evaluation.json",
)
xgb_step_eval = ProcessingStep(
name="XGBEval",
processor=xgb_script_eval,
inputs=[
ProcessingInput(
source=xgb_step_tuning.get_top_model_s3_uri(top_k=0, s3_bucket=model_bucket_key),
destination="/opt/ml/processing/model",
),
ProcessingInput(
source=step_process.properties.ProcessingOutputConfig.Outputs[
"test"
].S3Output.S3Uri,
destination="/opt/ml/processing/test",
),
],
outputs=[
ProcessingOutput(output_name="xgb_evaluation", source="/opt/ml/processing/evaluation"),
],
code=os.path.join(BASE_DIR, "xgb_evaluate.py"),
property_files=[xgb_evaluation_report],
)
xgb_model_metrics = ModelMetrics(
model_statistics=MetricsSource(
s3_uri="{}/xgb_evaluation.json".format(
xgb_step_eval.arguments["ProcessingOutputConfig"]["Outputs"][0]["S3Output"]["S3Uri"]
),
content_type="application/json",
)
)
dtree_model_metrics = ModelMetrics(
model_statistics=MetricsSource(
s3_uri="{}/dtree_evaluation.json".format(
dtree_step_eval.arguments["ProcessingOutputConfig"]["Outputs"][0]["S3Output"][
"S3Uri"
]
),
content_type="application/json",
)
)
xgb_eval_metrics = JsonGet(
step=xgb_step_eval,
property_file=xgb_evaluation_report,
json_path="regression_metrics.r2s.value", # This should follow the structure of your report_dict defined in the evaluate.py file.
)
dtree_eval_metrics = JsonGet(
step=dtree_step_eval,
property_file=dtree_evaluation_report,
json_path="regression_metrics.r2s.value", # This should follow the structure of your report_dict defined in the evaluate.py file.
)
# Register model step that will be conditionally executed
dtree_step_register = RegisterModel(
name="DTreeReg",
estimator=dtree_train,
model_data=dtree_step_tuning.get_top_model_s3_uri(top_k=0, s3_bucket=model_bucket_key),
content_types=["text/csv"],
response_types=["text/csv"],
inference_instances=["ml.t2.medium", "ml.m5.large"],
transform_instances=["ml.m5.large"],
model_package_group_name=model_package_group_name,
approval_status=model_approval_status,
model_metrics=dtree_model_metrics,
)
# Register model step that will be conditionally executed
xgb_step_register = RegisterModel(
name="XGBReg",
estimator=xgb_train,
model_data=xgb_step_tuning.get_top_model_s3_uri(top_k=0, s3_bucket=model_bucket_key),
content_types=["text/csv"],
response_types=["text/csv"],
inference_instances=["ml.t2.medium", "ml.m5.large"],
transform_instances=["ml.m5.large"],
model_package_group_name=model_package_group_name,
approval_status=model_approval_status,
model_metrics=xgb_model_metrics,
)
# Condition step for evaluating model quality and branching execution
cond_lte = ConditionGreaterThanOrEqualTo( # You can change the condition here
left=JsonGet(
step=dtree_step_eval,
property_file=dtree_evaluation_report,
json_path="regression_metrics.r2s.value", # This should follow the structure of your report_dict defined in the evaluate.py file.
),
right=JsonGet(
step=xgb_step_eval,
property_file=xgb_evaluation_report,
json_path="regression_metrics.r2s.value", # This should follow the structure of your report_dict defined in the evaluate.py file.
), # You can change the threshold here
)
step_cond = ConditionStep(
name="AccuracyCond",
conditions=[cond_lte],
if_steps=[dtree_step_register],
else_steps=[xgb_step_register],
)
create_date = time.strftime("%Y-%m-%d-%H-%M-%S")
# Pipeline instance
pipeline = Pipeline(
name=pipeline_name,
parameters=[
processing_instance_type,
processing_instance_count,
training_instance_type,
model_approval_status,
input_data
],
pipeline_experiment_config=PipelineExperimentConfig(
pipeline_name + "-" + create_date, "restate-{}".format(create_date)
),
steps=[
data_wrangler_step,
step_process,
dtree_step_tuning,
xgb_step_tuning,
dtree_step_eval,
xgb_step_eval,
step_cond,
],
sagemaker_session=sagemaker_session,
)
return pipeline
def test_pipeline_execution_processing_step_with_retry(
sagemaker_session,
smclient,
role,
sklearn_latest_version,
cpu_instance_type,
pipeline_name,
athena_dataset_definition,
):
instance_count = ParameterInteger(name="InstanceCount", default_value=2)
script_path = os.path.join(DATA_DIR, "dummy_script.py")
input_file_path = os.path.join(DATA_DIR, "dummy_input.txt")
inputs = [
ProcessingInput(source=input_file_path, destination="/opt/ml/processing/inputs/"),
ProcessingInput(dataset_definition=athena_dataset_definition),
]
sklearn_processor = SKLearnProcessor(
framework_version=sklearn_latest_version,
role=role,
instance_type=cpu_instance_type,
instance_count=instance_count,
command=["python3"],
sagemaker_session=sagemaker_session,
base_job_name="test-sklearn",
)
step_sklearn = ProcessingStep(
name="sklearn-process",
processor=sklearn_processor,
inputs=inputs,
code=script_path,
retry_policies=[
StepRetryPolicy(
exception_types=[
StepExceptionTypeEnum.SERVICE_FAULT,
StepExceptionTypeEnum.THROTTLING,
],
backoff_rate=2.0,
interval_seconds=30,
expire_after_mins=5,
),
SageMakerJobStepRetryPolicy(
exception_types=[SageMakerJobExceptionTypeEnum.CAPACITY_ERROR], max_attempts=10
),
],
)
pipeline = Pipeline(
name=pipeline_name,
parameters=[instance_count],
steps=[step_sklearn],
sagemaker_session=sagemaker_session,
)
try:
pipeline.create(role)
execution = pipeline.start(parameters={})
try:
execution.wait(delay=30, max_attempts=3)
except WaiterError:
pass
execution_steps = execution.list_steps()
assert len(execution_steps) == 1
assert execution_steps[0]["StepName"] == "sklearn-process"
# assert execution_steps[0]["AttemptCount"] >= 1
finally:
try:
pipeline.delete()
except Exception:
pass
def test_sklearn_with_all_customizations(sagemaker_session):
sklearn_processor = SKLearnProcessor(
framework_version="0.20.0",
role=ROLE,
command=["python3"],
instance_type="ml.m4.xlarge",
instance_count=1,
volume_size_in_gb=100,
volume_kms_key=None,
output_kms_key="arn:aws:kms:us-west-2:012345678901:key/kms-key",
max_runtime_in_seconds=3600,
base_job_name="my_sklearn_processor",
env={"my_env_variable": "my_env_variable_value"},
tags=[{
"Key": "my-tag",
"Value": "my-tag-value"
}],
network_config=NetworkConfig(
subnets=["my_subnet_id"],
security_group_ids=["my_security_group_id"],
enable_network_isolation=True,
),
sagemaker_session=sagemaker_session,
)
with patch("os.path.isdir", return_value=True):
sklearn_processor.run(
code="/local/path/to/sklearn_transformer.py",
inputs=[
ProcessingInput(
source="s3://path/to/my/dataset/census.csv",
destination="/container/path/",
input_name="my_dataset",
s3_data_type="S3Prefix",
s3_input_mode="File",
s3_data_distribution_type="FullyReplicated",
s3_compression_type="None",
)
],
outputs=[
ProcessingOutput(
source="/container/path/",
destination="s3://uri/",
output_name="my_output",
s3_upload_mode="EndOfJob",
)
],
arguments=["--drop-columns", "'SelfEmployed'"],
wait=True,
logs=False,
job_name="my_job_name",
experiment_config={"ExperimentName": "AnExperiment"},
)
expected_args = {
"inputs": [
{
"InputName": "my_dataset",
"S3Input": {
"S3Uri": "s3://path/to/my/dataset/census.csv",
"LocalPath": "/container/path/",
"S3DataType": "S3Prefix",
"S3InputMode": "File",
"S3DataDistributionType": "FullyReplicated",
"S3CompressionType": "None",
},
},
{
"InputName": "code",
"S3Input": {
"S3Uri": "mocked_s3_uri_from_upload_data",
"LocalPath": "/opt/ml/processing/input/code",
"S3DataType": "S3Prefix",
"S3InputMode": "File",
"S3DataDistributionType": "FullyReplicated",
"S3CompressionType": "None",
},
},
],
"output_config": {
"Outputs": [{
"OutputName": "my_output",
"S3Output": {
"S3Uri": "s3://uri/",
"LocalPath": "/container/path/",
"S3UploadMode": "EndOfJob",
},
}],
"KmsKeyId":
"arn:aws:kms:us-west-2:012345678901:key/kms-key",
},
"job_name":
sklearn_processor._current_job_name,
"resources": {
"ClusterConfig": {
"InstanceType": "ml.m4.xlarge",
"InstanceCount": 1,
"VolumeSizeInGB": 100,
}
},
"stopping_condition": {
"MaxRuntimeInSeconds": 3600
},
"app_specification": {
"ImageUri":
"246618743249.dkr.ecr.us-west-2.amazonaws.com/sagemaker-scikit-learn:0.20.0-cpu-py3",
"ContainerArguments": ["--drop-columns", "'SelfEmployed'"],
"ContainerEntrypoint": [
"python3",
"/opt/ml/processing/input/code/sklearn_transformer.py",
],
},
"environment": {
"my_env_variable": "my_env_variable_value"
},
"network_config": {
"EnableNetworkIsolation": True,
"VpcConfig": {
"SecurityGroupIds": ["my_security_group_id"],
"Subnets": ["my_subnet_id"],
},
},
"role_arn":
ROLE,
"tags": [{
"Key": "my-tag",
"Value": "my-tag-value"
}],
"experiment_config": {
"ExperimentName": "AnExperiment"
},
}
sagemaker_session.process.assert_called_with(**expected_args)
def get_pipeline(
region,
security_group_ids,
subnets,
processing_role=None,
training_role=None,
data_bucket=None,
model_bucket=None,
model_package_group_name="AbalonePackageGroup",
pipeline_name="AbalonePipeline",
base_job_prefix="Abalone",
):
"""Gets a SageMaker ML Pipeline instance working with on abalone data.
Args:
region: AWS region to create and run the pipeline.
processing_role: IAM role to create and run processing steps
training_role: IAM role to create and run training steps
data_bucket: the bucket to use for storing the artifacts
Returns:
an instance of a pipeline
"""
sagemaker_session = get_session(region, data_bucket)
if processing_role is None:
processing_role = sagemaker.session.get_execution_role(sagemaker_session)
if training_role is None:
training_role = sagemaker.session.get_execution_role(sagemaker_session)
if model_bucket is None:
model_bucket = sagemaker_session.default_bucket()
print(f"Creating the pipeline '{pipeline_name}':")
print(f"Parameters:{region}\n{security_group_ids}\n{subnets}\n{processing_role}\n\
{training_role}\n{data_bucket}\n{model_bucket}\n{model_package_group_name}\n\
{pipeline_name}\n{base_job_prefix}")
# parameters for pipeline execution
processing_instance_count = ParameterInteger(name="ProcessingInstanceCount", default_value=1)
processing_instance_type = ParameterString(
name="ProcessingInstanceType", default_value="ml.m5.xlarge"
)
training_instance_type = ParameterString(
name="TrainingInstanceType", default_value="ml.m5.xlarge"
)
model_approval_status = ParameterString(
name="ModelApprovalStatus", default_value="PendingManualApproval"
)
input_data = ParameterString(
name="InputDataUrl",
default_value=f"s3://{sagemaker_session.default_bucket()}/datasets/abalone-dataset.csv",
)
# configure network for encryption, network isolation and VPC configuration
# Since the preprocessor job takes the data from S3, enable_network_isolation must be set to False
# see https://github.com/aws/amazon-sagemaker-examples/issues/1689
network_config = NetworkConfig(
enable_network_isolation=False,
security_group_ids=security_group_ids.split(","),
subnets=subnets.split(","),
encrypt_inter_container_traffic=True)
# processing step for feature engineering
sklearn_processor = SKLearnProcessor(
framework_version="0.23-1",
instance_type=processing_instance_type,
instance_count=processing_instance_count,
base_job_name=f"{base_job_prefix}/sklearn-abalone-preprocess",
sagemaker_session=sagemaker_session,
role=processing_role,
network_config=network_config
)
step_process = ProcessingStep(
name="PreprocessAbaloneData",
processor=sklearn_processor,
outputs=[
ProcessingOutput(output_name="train", source="/opt/ml/processing/train"),
ProcessingOutput(output_name="validation", source="/opt/ml/processing/validation"),
ProcessingOutput(output_name="test", source="/opt/ml/processing/test"),
],
code=os.path.join(BASE_DIR, "preprocess.py"),
job_arguments=["--input-data", input_data],
)
# training step for generating model artifacts
model_path = f"s3://{model_bucket}/{base_job_prefix}/AbaloneTrain"
image_uri = sagemaker.image_uris.retrieve(
framework="xgboost",
region=region,
version="1.0-1",
py_version="py3",
instance_type=training_instance_type,
)
xgb_train = Estimator(
image_uri=image_uri,
instance_type=training_instance_type,
instance_count=1,
output_path=model_path,
base_job_name=f"{base_job_prefix}/abalone-train",
sagemaker_session=sagemaker_session,
role=training_role,
subnets=network_config.subnets,
security_group_ids=network_config.security_group_ids,
encrypt_inter_container_traffic=True,
enable_network_isolation=False
)
xgb_train.set_hyperparameters(
objective="reg:linear",
num_round=50,
max_depth=5,
eta=0.2,
gamma=4,
min_child_weight=6,
subsample=0.7,
silent=0,
)
step_train = TrainingStep(
name="TrainAbaloneModel",
estimator=xgb_train,
inputs={
"train": TrainingInput(
s3_data=step_process.properties.ProcessingOutputConfig.Outputs[
"train"
].S3Output.S3Uri,
content_type="text/csv",
),
"validation": TrainingInput(
s3_data=step_process.properties.ProcessingOutputConfig.Outputs[
"validation"
].S3Output.S3Uri,
content_type="text/csv",
),
},
)
# processing step for evaluation
script_eval = ScriptProcessor(
image_uri=image_uri,
command=["python3"],
instance_type=processing_instance_type,
instance_count=1,
base_job_name=f"{base_job_prefix}/script-abalone-eval",
sagemaker_session=sagemaker_session,
role=processing_role,
network_config=network_config
)
evaluation_report = PropertyFile(
name="AbaloneEvaluationReport",
output_name="evaluation",
path="evaluation.json",
)
step_eval = ProcessingStep(
name="EvaluateAbaloneModel",
processor=script_eval,
inputs=[
ProcessingInput(
source=step_train.properties.ModelArtifacts.S3ModelArtifacts,
destination="/opt/ml/processing/model",
),
ProcessingInput(
source=step_process.properties.ProcessingOutputConfig.Outputs[
"test"
].S3Output.S3Uri,
destination="/opt/ml/processing/test",
),
],
outputs=[
ProcessingOutput(output_name="evaluation", source="/opt/ml/processing/evaluation"),
],
code=os.path.join(BASE_DIR, "evaluate.py"),
property_files=[evaluation_report],
)
# register model step that will be conditionally executed
model_metrics = ModelMetrics(
model_statistics=MetricsSource(
s3_uri="{}/evaluation.json".format(
step_eval.arguments["ProcessingOutputConfig"]["Outputs"][0]["S3Output"]["S3Uri"]
),
content_type="application/json"
)
)
"""
There is a bug in RegisterModel implementation
The RegisterModel step is implemented in the SDK as two steps, a _RepackModelStep and a _RegisterModelStep.
The _RepackModelStep runs a SKLearn training step in order to repack the model.tar.gz to include any custom inference code in the archive.
The _RegisterModelStep then registers the repacked model.
The problem is that the _RepackModelStep does not propagate VPC configuration from the Estimator object:
https://github.com/aws/sagemaker-python-sdk/blob/cdb633b3ab02398c3b77f5ecd2c03cdf41049c78/src/sagemaker/workflow/_utils.py#L88
This cause the AccessDenied exception because repacker cannot access S3 bucket (all access which is not via VPC endpoint is bloked by the bucket policy)
The issue is opened against SageMaker python SDK: https://github.com/aws/sagemaker-python-sdk/issues/2302
"""
vpc_config = {
"Subnets":network_config.subnets,
"SecurityGroupIds":network_config.security_group_ids
}
step_register = RegisterModel(
name="RegisterAbaloneModel",
estimator=xgb_train,
model_data=step_train.properties.ModelArtifacts.S3ModelArtifacts,
content_types=["text/csv"],
response_types=["text/csv"],
inference_instances=["ml.t2.medium", "ml.m5.large"],
transform_instances=["ml.m5.large"],
model_package_group_name=model_package_group_name,
approval_status=model_approval_status,
model_metrics=model_metrics,
vpc_config_override=vpc_config
)
# condition step for evaluating model quality and branching execution
cond_lte = ConditionLessThanOrEqualTo(
left=JsonGet(
step=step_eval,
property_file=evaluation_report,
json_path="regression_metrics.mse.value"
),
right=6.0,
)
step_cond = ConditionStep(
name="CheckMSEAbaloneEvaluation",
conditions=[cond_lte],
if_steps=[step_register],
else_steps=[],
)
# pipeline instance
pipeline = Pipeline(
name=pipeline_name,
parameters=[
processing_instance_type,
processing_instance_count,
training_instance_type,
model_approval_status,
input_data,
],
steps=[step_process, step_train, step_eval, step_cond],
sagemaker_session=sagemaker_session,
)
return pipeline
def get_pipeline(region, role, default_bucket, pipeline_name,
model_package_group_name, base_job_prefix):
"""Gets a SageMaker ML Pipeline instance working with BERT.
Args:
region: AWS region to create and run the pipeline.
role: IAM role to create and run steps and pipeline.
default_bucket: the bucket to use for storing the artifacts
pipeline_name: name of this pipeline
model_package_group_name: model package group
base_job_prefix: prefic of the job name
Returns:
an instance of a pipeline
"""
sm = boto3.Session().client(service_name="sagemaker", region_name=region)
input_data = ParameterString(
name="InputDataUrl",
default_value="s3://{}/amazon-reviews-pds/tsv/".format(bucket),
)
processing_instance_count = ParameterInteger(
name="ProcessingInstanceCount", default_value=1)
processing_instance_type = ParameterString(name="ProcessingInstanceType",
default_value="ml.c5.2xlarge")
max_seq_length = ParameterInteger(
name="MaxSeqLength",
default_value=64,
)
balance_dataset = ParameterString(
name="BalanceDataset",
default_value="True",
)
train_split_percentage = ParameterFloat(
name="TrainSplitPercentage",
default_value=0.90,
)
validation_split_percentage = ParameterFloat(
name="ValidationSplitPercentage",
default_value=0.05,
)
test_split_percentage = ParameterFloat(
name="TestSplitPercentage",
default_value=0.05,
)
feature_store_offline_prefix = ParameterString(
name="FeatureStoreOfflinePrefix",
default_value="reviews-feature-store-" + str(timestamp),
)
feature_group_name = ParameterString(
name="FeatureGroupName",
default_value="reviews-feature-group-" + str(timestamp))
train_instance_type = ParameterString(name="TrainInstanceType",
default_value="ml.c5.9xlarge")
train_instance_count = ParameterInteger(name="TrainInstanceCount",
default_value=1)
#########################
# PROCESSING STEP
#########################
processor = SKLearnProcessor(
framework_version="0.23-1",
role=role,
instance_type=processing_instance_type,
instance_count=processing_instance_count,
env={"AWS_DEFAULT_REGION": region},
max_runtime_in_seconds=7200,
)
processing_inputs = [
ProcessingInput(
input_name="raw-input-data",
source=input_data,
destination="/opt/ml/processing/input/data/",
s3_data_distribution_type="ShardedByS3Key",
)
]
processing_outputs = [
ProcessingOutput(
output_name="bert-train",
s3_upload_mode="EndOfJob",
source="/opt/ml/processing/output/bert/train",
),
ProcessingOutput(
output_name="bert-validation",
s3_upload_mode="EndOfJob",
source="/opt/ml/processing/output/bert/validation",
),
ProcessingOutput(
output_name="bert-test",
s3_upload_mode="EndOfJob",
source="/opt/ml/processing/output/bert/test",
),
]
# TODO: Figure out why the Parameter's are not resolving properly to their native type when user here.
# We shouldn't be using `default_value`
processing_step = ProcessingStep(
name="Processing",
processor=processor,
inputs=processing_inputs,
outputs=processing_outputs,
job_arguments=[
"--train-split-percentage",
str(train_split_percentage.default_value),
"--validation-split-percentage",
str(validation_split_percentage.default_value),
"--test-split-percentage",
str(test_split_percentage.default_value),
"--max-seq-length",
str(max_seq_length.default_value),
"--balance-dataset",
str(balance_dataset.default_value),
"--feature-store-offline-prefix",
str(feature_store_offline_prefix.default_value),
"--feature-group-name",
str(feature_group_name.default_value),
],
code=os.path.join(BASE_DIR,
"preprocess-scikit-text-to-bert-feature-store.py"),
)
#########################
# TRAINING STEP
#########################
epochs = ParameterInteger(name="Epochs", default_value=1)
learning_rate = ParameterFloat(name="LearningRate", default_value=0.00001)
epsilon = ParameterFloat(name="Epsilon", default_value=0.00000001)
train_batch_size = ParameterInteger(name="TrainBatchSize",
default_value=128)
validation_batch_size = ParameterInteger(name="ValidationBatchSize",
default_value=128)
test_batch_size = ParameterInteger(name="TestBatchSize", default_value=128)
train_steps_per_epoch = ParameterInteger(name="TrainStepsPerEpoch",
default_value=50)
validation_steps = ParameterInteger(name="ValidationSteps",
default_value=50)
test_steps = ParameterInteger(name="TestSteps", default_value=50)
train_volume_size = ParameterInteger(name="TrainVolumeSize",
default_value=1024)
use_xla = ParameterString(
name="UseXLA",
default_value="True",
)
use_amp = ParameterString(
name="UseAMP",
default_value="True",
)
freeze_bert_layer = ParameterString(
name="FreezeBERTLayer",
default_value="False",
)
enable_sagemaker_debugger = ParameterString(
name="EnableSageMakerDebugger",
default_value="False",
)
enable_checkpointing = ParameterString(
name="EnableCheckpointing",
default_value="False",
)
enable_tensorboard = ParameterString(
name="EnableTensorboard",
default_value="False",
)
input_mode = ParameterString(
name="InputMode",
default_value="File",
)
run_validation = ParameterString(
name="RunValidation",
default_value="True",
)
run_test = ParameterString(
name="RunTest",
default_value="False",
)
run_sample_predictions = ParameterString(
name="RunSamplePredictions",
default_value="False",
)
metrics_definitions = [
{
"Name": "train:loss",
"Regex": "loss: ([0-9\\.]+)"
},
{
"Name": "train:accuracy",
"Regex": "accuracy: ([0-9\\.]+)"
},
{
"Name": "validation:loss",
"Regex": "val_loss: ([0-9\\.]+)"
},
{
"Name": "validation:accuracy",
"Regex": "val_accuracy: ([0-9\\.]+)"
},
]
train_src = os.path.join(BASE_DIR, "src")
model_path = f"s3://{default_bucket}/{base_job_prefix}/output/model"
estimator = TensorFlow(
entry_point="tf_bert_reviews.py",
source_dir=BASE_DIR,
role=role,
output_path=model_path,
instance_count=train_instance_count,
instance_type=train_instance_type,
volume_size=train_volume_size,
py_version="py37",
framework_version="2.3.1",
hyperparameters={
"epochs": epochs,
"learning_rate": learning_rate,
"epsilon": epsilon,
"train_batch_size": train_batch_size,
"validation_batch_size": validation_batch_size,
"test_batch_size": test_batch_size,
"train_steps_per_epoch": train_steps_per_epoch,
"validation_steps": validation_steps,
"test_steps": test_steps,
"use_xla": use_xla,
"use_amp": use_amp,
"max_seq_length": max_seq_length,
"freeze_bert_layer": freeze_bert_layer,
"enable_sagemaker_debugger": enable_sagemaker_debugger,
"enable_checkpointing": enable_checkpointing,
"enable_tensorboard": enable_tensorboard,
"run_validation": run_validation,
"run_test": run_test,
"run_sample_predictions": run_sample_predictions,
},
input_mode=input_mode,
metric_definitions=metrics_definitions,
# max_run=7200 # max 2 hours * 60 minutes seconds per hour * 60 seconds per minute
)
training_step = TrainingStep(
name="Train",
estimator=estimator,
inputs={
"train":
TrainingInput(
s3_data=processing_step.properties.ProcessingOutputConfig.
Outputs["bert-train"].S3Output.S3Uri,
content_type="text/csv",
),
"validation":
TrainingInput(
s3_data=processing_step.properties.ProcessingOutputConfig.
Outputs["bert-validation"].S3Output.S3Uri,
content_type="text/csv",
),
"test":
TrainingInput(
s3_data=processing_step.properties.ProcessingOutputConfig.
Outputs["bert-test"].S3Output.S3Uri,
content_type="text/csv",
),
},
)
#########################
# EVALUATION STEP
#########################
evaluation_processor = SKLearnProcessor(
framework_version="0.23-1",
role=role,
instance_type=processing_instance_type,
instance_count=processing_instance_count,
env={"AWS_DEFAULT_REGION": region},
max_runtime_in_seconds=7200,
)
evaluation_report = PropertyFile(name="EvaluationReport",
output_name="metrics",
path="evaluation.json")
evaluation_step = ProcessingStep(
name="EvaluateModel",
processor=evaluation_processor,
code=os.path.join(BASE_DIR, "evaluate_model_metrics.py"),
inputs=[
ProcessingInput(
source=training_step.properties.ModelArtifacts.
S3ModelArtifacts,
destination="/opt/ml/processing/input/model",
),
ProcessingInput(
source=processing_step.properties.
ProcessingInputs["raw-input-data"].S3Input.S3Uri,
destination="/opt/ml/processing/input/data",
),
],
outputs=[
ProcessingOutput(output_name="metrics",
s3_upload_mode="EndOfJob",
source="/opt/ml/processing/output/metrics/"),
],
job_arguments=[
"--max-seq-length",
str(max_seq_length.default_value),
],
property_files=[evaluation_report
], # these cause deserialization issues
)
model_metrics = ModelMetrics(model_statistics=MetricsSource(
s3_uri="{}/evaluation.json".format(
evaluation_step.arguments["ProcessingOutputConfig"]["Outputs"][0]
["S3Output"]["S3Uri"]),
content_type="application/json",
))
#########################
## REGISTER TRAINED MODEL STEP
#########################
model_approval_status = ParameterString(
name="ModelApprovalStatus", default_value="PendingManualApproval")
deploy_instance_type = ParameterString(name="DeployInstanceType",
default_value="ml.m5.4xlarge")
deploy_instance_count = ParameterInteger(name="DeployInstanceCount",
default_value=1)
inference_image_uri = sagemaker.image_uris.retrieve(
framework="tensorflow",
region=region,
version="2.3.1",
py_version="py37",
instance_type=deploy_instance_type,
image_scope="inference",
)
print(inference_image_uri)
register_step = RegisterModel(
name="RegisterModel",
estimator=estimator,
image_uri=
inference_image_uri, # we have to specify, by default it's using training image
model_data=training_step.properties.ModelArtifacts.S3ModelArtifacts,
content_types=["text/csv"],
response_types=["text/csv"],
inference_instances=[
deploy_instance_type
], # The JSON spec must be within these instance types or we will see "Instance Type Not Allowed" Exception
transform_instances=[deploy_instance_type],
model_package_group_name=model_package_group_name,
approval_status=model_approval_status,
)
#########################
## CREATE MODEL FOR DEPLOYMENT STEP
#########################
model = Model(
image_uri=inference_image_uri,
model_data=training_step.properties.ModelArtifacts.S3ModelArtifacts,
sagemaker_session=sess,
role=role,
)
create_inputs = CreateModelInput(instance_type=deploy_instance_type, )
create_step = CreateModelStep(
name="CreateModel",
model=model,
inputs=create_inputs,
)
#########################
## CONDITION STEP: EVALUATE THE MODEL
#########################
min_accuracy_value = ParameterFloat(name="MinAccuracyValue",
default_value=0.01)
minimum_accuracy_condition = ConditionGreaterThanOrEqualTo(
left=JsonGet(
step=evaluation_step,
property_file=evaluation_report,
json_path="metrics.accuracy.value",
),
right=min_accuracy_value, # accuracy
)
minimum_accuracy_condition_step = ConditionStep(
name="AccuracyCondition",
conditions=[minimum_accuracy_condition],
if_steps=[register_step,
create_step], # success, continue with model registration
else_steps=[], # fail, end the pipeline
)
#########################
## CREATE PIPELINE
#########################
pipeline = Pipeline(
name=pipeline_name,
parameters=[
input_data,
processing_instance_count,
processing_instance_type,
max_seq_length,
balance_dataset,
train_split_percentage,
validation_split_percentage,
test_split_percentage,
feature_store_offline_prefix,
feature_group_name,
train_instance_type,
train_instance_count,
epochs,
learning_rate,
epsilon,
train_batch_size,
validation_batch_size,
test_batch_size,
train_steps_per_epoch,
validation_steps,
test_steps,
train_volume_size,
use_xla,
use_amp,
freeze_bert_layer,
enable_sagemaker_debugger,
enable_checkpointing,
enable_tensorboard,
input_mode,
run_validation,
run_test,
run_sample_predictions,
min_accuracy_value,
model_approval_status,
deploy_instance_type,
deploy_instance_count,
],
steps=[
processing_step, training_step, evaluation_step,
minimum_accuracy_condition_step
],
sagemaker_session=sess,
)
#########################
## RETURN PIPELINE
#########################
return pipeline
# 3. You should have AWS credentials configured on your local machine
# in order to be able to pull the docker image from ECR.
########################################################################################################################
from sagemaker.local import LocalSession
from sagemaker.processing import ProcessingInput, ProcessingOutput
from sagemaker.sklearn.processing import SKLearnProcessor
sagemaker_session = LocalSession()
sagemaker_session.config = {'local': {'local_code': True}}
# For local training a dummy role will be sufficient
role = 'arn:aws:iam::111111111111:role/service-role/AmazonSageMaker-ExecutionRole-20200101T000001'
processor = SKLearnProcessor(framework_version='0.20.0',
instance_count=1,
instance_type='local',
role=role)
print('Starting processing job.')
print('Note: if launching for the first time in local mode, container image download might take a few minutes to complete.')
processor.run(code='processing_script.py',
inputs=[ProcessingInput(
source='./input_data/',
destination='/opt/ml/processing/input_data/')],
outputs=[ProcessingOutput(
output_name='word_count_data',
source='/opt/ml/processing/processed_data/')],
arguments=['job-type', 'word-count']
)
preprocessing_job_description = processor.jobs[-1].describe()
def get_pipeline(
region,
role=None,
default_bucket=None,
model_package_group_name="TestPackageGroup",
pipeline_name="TestPipeline",
base_job_prefix="Test",
):
"""Gets a SageMaker ML Pipeline instance working with on abalone data.
Args:
region: AWS region to create and run the pipeline.
role: IAM role to create and run steps and pipeline.
default_bucket: the bucket to use for storing the artifacts
Returns:
an instance of a pipeline
"""
sagemaker_session = get_session(region, default_bucket)
if role is None:
role = sagemaker.session.get_execution_role(sagemaker_session)
# parameters for pipeline execution
processing_instance_count = ParameterInteger(
name="ProcessingInstanceCount", default_value=1)
processing_instance_type = ParameterString(name="ProcessingInstanceType",
default_value="ml.m5.xlarge")
training_instance_type = ParameterString(name="TrainingInstanceType",
default_value="ml.m5.xlarge")
model_approval_status = ParameterString(
name="ModelApprovalStatus", default_value="PendingManualApproval")
input_data = ParameterString(
name="InputDataUrl",
default_value=
f"s3://sagemaker-servicecatalog-seedcode-{region}/dataset/abalone-dataset.csv",
)
# processing step for feature engineering
sklearn_processor = SKLearnProcessor(
framework_version="0.23-1",
instance_type=processing_instance_type,
instance_count=processing_instance_count,
base_job_name=f"{base_job_prefix}/sklearn-test-preprocess",
sagemaker_session=sagemaker_session,
role=role,
)
step_process = ProcessingStep(
name="PreprocessTestData",
processor=sklearn_processor,
outputs=[
ProcessingOutput(output_name="train",
source="/opt/ml/processing/train"),
ProcessingOutput(output_name="validation",
source="/opt/ml/processing/validation"),
ProcessingOutput(output_name="test",
source="/opt/ml/processing/test"),
],
code=os.path.join(BASE_DIR, "preprocess.py"),
job_arguments=["--input-data", input_data],
)
# training step for generating model artifacts
model_path = f"s3://{sagemaker_session.default_bucket()}/{base_job_prefix}/TestTrain"
image_uri = sagemaker.image_uris.retrieve(
framework="xgboost",
region=region,
version="1.0-1",
py_version="py3",
instance_type=training_instance_type,
)
xgb_train = Estimator(
image_uri=image_uri,
instance_type=training_instance_type,
instance_count=1,
output_path=model_path,
base_job_name=f"{base_job_prefix}/test-train",
sagemaker_session=sagemaker_session,
role=role,
)
xgb_train.set_hyperparameters(
objective="reg:linear",
num_round=50,
max_depth=5,
eta=0.2,
gamma=4,
min_child_weight=6,
subsample=0.7,
silent=0,
)
step_train = TrainingStep(
name="TrainTestModel",
estimator=xgb_train,
inputs={
"train":
TrainingInput(
s3_data=step_process.properties.ProcessingOutputConfig.
Outputs["train"].S3Output.S3Uri,
content_type="text/csv",
),
"validation":
TrainingInput(
s3_data=step_process.properties.ProcessingOutputConfig.
Outputs["validation"].S3Output.S3Uri,
content_type="text/csv",
),
},
)
# processing step for evaluation
script_eval = ScriptProcessor(
image_uri=image_uri,
command=["python3"],
instance_type=processing_instance_type,
instance_count=1,
base_job_name=f"{base_job_prefix}/script-test-eval",
sagemaker_session=sagemaker_session,
role=role,
)
evaluation_report = PropertyFile(
name="TestEvaluationReport",
output_name="evaluation",
path="evaluation.json",
)
step_eval = ProcessingStep(
name="EvaluateTestModel",
processor=script_eval,
inputs=[
ProcessingInput(
source=step_train.properties.ModelArtifacts.S3ModelArtifacts,
destination="/opt/ml/processing/model",
),
ProcessingInput(
source=step_process.properties.ProcessingOutputConfig.
Outputs["test"].S3Output.S3Uri,
destination="/opt/ml/processing/test",
),
],
outputs=[
ProcessingOutput(output_name="evaluation",
source="/opt/ml/processing/evaluation"),
],
code=os.path.join(BASE_DIR, "evaluate.py"),
property_files=[evaluation_report],
)
# register model step that will be conditionally executed
model_metrics = ModelMetrics(
model_statistics=MetricsSource(s3_uri="{}/evaluation.json".format(
step_eval.arguments["ProcessingOutputConfig"]["Outputs"][0]
["S3Output"]["S3Uri"]),
content_type="application/json"))
step_register = RegisterModel(
name="RegisterTestModel",
estimator=xgb_train,
model_data=step_train.properties.ModelArtifacts.S3ModelArtifacts,
content_types=["text/csv"],
response_types=["text/csv"],
inference_instances=["ml.t2.medium", "ml.m5.large"],
transform_instances=["ml.m5.large"],
model_package_group_name=model_package_group_name,
approval_status=model_approval_status,
model_metrics=model_metrics,
)
# condition step for evaluating model quality and branching execution
cond_lte = ConditionLessThanOrEqualTo(
left=JsonGet(step=step_eval,
property_file=evaluation_report,
json_path="regression_metrics.mse.value"),
right=6.0,
)
step_cond = ConditionStep(
name="CheckMSETestEvaluation",
conditions=[cond_lte],
if_steps=[step_register],
else_steps=[],
)
# pipeline instance
pipeline = Pipeline(
name=pipeline_name,
parameters=[
processing_instance_type,
processing_instance_count,
training_instance_type,
model_approval_status,
input_data,
],
steps=[step_process, step_train, step_eval, step_cond],
sagemaker_session=sagemaker_session,
)
return pipeline
def define_inference_pipeline(
sm_role,
workflow_execution_role,
inference_pipeline_name,
return_yaml=True,
dump_yaml_file="templates/sagemaker_inference_pipeline.yaml",
kms_key_id=None,
):
"""
Return YAML definition of the training pipeline, which consists of multiple
Amazon StepFunction steps
sm_role: ARN of the SageMaker execution role
workflow_execution_role: ARN of the StepFunction execution role
return_yaml: Return YAML representation or not, if False,
it returns an instance of `stepfunctions.workflow.WorkflowObject`
dump_yaml_file: If not None, a YAML file will be generated at this file location
"""
# Pass required parameters dynamically for each execution using placeholders.
execution_input = ExecutionInput(
schema={
"InputDataURL": str,
"PreprocessingJobName": str,
"InferenceJobName": str,
"ProcModelS3": str,
"PreprocessingCodeURL": str,
"InferenceCodeURL": str,
"ModelS3": str,
"PreprocessedTrainDataURL": str,
"PreprocessedTestDataURL": str,
"OutputPathURL": str,
})
"""
Create Preprocessing Model from model artifact.
"""
# sagemaker_session = sagemaker.Session()
sklearn_processor = SKLearnProcessor(
framework_version="0.20.0",
role=sm_role,
instance_type="ml.m5.xlarge",
instance_count=1,
max_runtime_in_seconds=1200,
)
# Create ProcessingInputs and ProcessingOutputs objects for Inputs and
# Outputs respectively for the SageMaker Processing Job
inputs = [
ProcessingInput(
source=execution_input["InputDataURL"],
destination="/opt/ml/processing/input",
input_name="input-1",
),
ProcessingInput(
source=execution_input["PreprocessingCodeURL"],
destination="/opt/ml/processing/input/code",
input_name="code",
),
ProcessingInput(
source=execution_input["ProcModelS3"],
destination="/opt/ml/processing/model",
input_name="proc_model",
),
]
outputs = [
ProcessingOutput(
source="/opt/ml/processing/test",
destination=execution_input["PreprocessedTestDataURL"],
output_name="test_data",
),
]
processing_step = ProcessingStep(
"SageMaker pre-processing step",
processor=sklearn_processor,
job_name=execution_input["PreprocessingJobName"],
inputs=inputs,
outputs=outputs,
container_arguments=["--mode", "infer"],
container_entrypoint=[
"python3",
"/opt/ml/processing/input/code/preprocessing.py",
],
kms_key_id=kms_key_id,
)
"""
Create inference with sklearn processing step.
Inputs are the preprocessed data S3 URL, the inference code S3 URL, and
the model S3 URL. Output is the inferred data.
"""
sklearn_processor2 = SKLearnProcessor(
framework_version="0.20.0",
role=sm_role,
instance_type="ml.m5.xlarge",
instance_count=1,
max_runtime_in_seconds=1200,
)
inputs = [
ProcessingInput(
source=execution_input["PreprocessedTestDataURL"],
destination="/opt/ml/processing/input",
input_name="input-1",
),
ProcessingInput(
source=execution_input["InferenceCodeURL"],
destination="/opt/ml/processing/input/code",
input_name="code",
),
ProcessingInput(
source=execution_input["ModelS3"],
destination="/opt/ml/processing/model",
input_name="model",
),
]
outputs = [
ProcessingOutput(
source="/opt/ml/processing/test",
destination=execution_input["OutputPathURL"],
output_name="test_data",
),
]
inference_step = ProcessingStep(
"SageMaker inference step",
processor=sklearn_processor2,
job_name=execution_input["InferenceJobName"],
inputs=inputs,
outputs=outputs,
container_entrypoint=[
"python3",
"/opt/ml/processing/input/code/inference.py",
],
kms_key_id=kms_key_id,
)
# Create Fail state to mark the workflow failed in case any of the steps fail.
failed_state_sagemaker_processing_failure = stepfunctions.steps.states.Fail(
"ML Workflow failed", cause="SageMakerProcessingJobFailed")
# Add the Error handling in the workflow
catch_state_processing = stepfunctions.steps.states.Catch(
error_equals=["States.TaskFailed"],
next_step=failed_state_sagemaker_processing_failure,
)
processing_step.add_catch(catch_state_processing)
inference_step.add_catch(catch_state_processing)
# Create the Workflow
workflow_graph = Chain([processing_step, inference_step])
inference_pipeline = Workflow(
name=inference_pipeline_name,
definition=workflow_graph,
role=workflow_execution_role,
)
return inference_pipeline
