def test_auto_ml_attach(sagemaker_session):
expected_default_input_config = [{
"DataSource": {
"S3DataSource": {
"S3DataType":
"S3Prefix",
"S3Uri":
"s3://{}/{}/input/iris_training.csv".format(
sagemaker_session.default_bucket(), PREFIX),
}
},
"TargetAttributeName":
TARGET_ATTRIBUTE_NAME,
}]
expected_default_output_config = {
"S3OutputPath": "s3://{}/".format(sagemaker_session.default_bucket())
}
auto_ml_utils.create_auto_ml_job_if_not_exist(sagemaker_session)
attached_automl_job = AutoML.attach(auto_ml_job_name=AUTO_ML_JOB_NAME,
sagemaker_session=sagemaker_session)
attached_desc = attached_automl_job.describe_auto_ml_job()
assert attached_desc["AutoMLJobName"] == AUTO_ML_JOB_NAME
assert attached_desc["AutoMLJobStatus"] == "Completed"
assert isinstance(attached_desc["BestCandidate"], dict)
assert attached_desc["InputDataConfig"] == expected_default_input_config
assert attached_desc["AutoMLJobConfig"] == EXPECTED_DEFAULT_JOB_CONFIG
assert attached_desc["OutputDataConfig"] == expected_default_output_config
def test_candidate_estimator_default_rerun_and_deploy(sagemaker_session,
cpu_instance_type):
auto_ml_utils.create_auto_ml_job_if_not_exist(sagemaker_session)
auto_ml = AutoML(role=ROLE,
target_attribute_name=TARGET_ATTRIBUTE_NAME,
sagemaker_session=sagemaker_session)
candidates = auto_ml.list_candidates(job_name=AUTO_ML_JOB_NAME)
candidate = candidates[1]
candidate_estimator = CandidateEstimator(candidate, sagemaker_session)
inputs = sagemaker_session.upload_data(path=TEST_DATA,
key_prefix=PREFIX + "/input")
endpoint_name = unique_name_from_base(
"sagemaker-auto-ml-rerun-candidate-test")
with timeout(minutes=AUTO_ML_DEFAULT_TIMEMOUT_MINUTES):
candidate_estimator.fit(inputs)
auto_ml.deploy(
initial_instance_count=INSTANCE_COUNT,
instance_type=cpu_instance_type,
candidate=candidate,
endpoint_name=endpoint_name,
)
endpoint_status = sagemaker_session.sagemaker_client.describe_endpoint(
EndpointName=endpoint_name)["EndpointStatus"]
assert endpoint_status == "InService"
sagemaker_session.sagemaker_client.delete_endpoint(
EndpointName=endpoint_name)
def test_create_model_best_candidate(sagemaker_session, cpu_instance_type):
auto_ml_utils.create_auto_ml_job_if_not_exist(sagemaker_session)
auto_ml = AutoML.attach(auto_ml_job_name=AUTO_ML_JOB_NAME, sagemaker_session=sagemaker_session)
best_candidate = auto_ml.best_candidate()
with timeout(minutes=5):
pipeline_model = auto_ml.create_model(
name=DEFAULT_MODEL_NAME,
candidate=best_candidate,
sagemaker_session=sagemaker_session,
vpc_config=None,
enable_network_isolation=False,
model_kms_key=None,
predictor_cls=None,
)
inputs = sagemaker_session.upload_data(
path=TRANSFORM_DATA, key_prefix=PREFIX + "/transform_input"
)
pipeline_model.transformer(
instance_count=1,
instance_type=cpu_instance_type,
assemble_with="Line",
output_path="s3://{}/{}".format(sagemaker_session.default_bucket(), "transform_test"),
accept="text/csv",
).transform(data=inputs, content_type="text/csv", split_type="Line", join_source="Input")
def test_auto_ml_describe_auto_ml_job(sagemaker_session):
expected_default_input_config = [{
"DataSource": {
"S3DataSource": {
"S3DataType":
"S3Prefix",
"S3Uri":
"s3://{}/{}/input/iris_training.csv".format(
sagemaker_session.default_bucket(), PREFIX),
}
},
"TargetAttributeName": TARGET_ATTRIBUTE_NAME,
"ContentType": "text/csv;header=present",
"ChannelType": "training",
}]
expected_default_output_config = {
"S3OutputPath": "s3://{}/".format(sagemaker_session.default_bucket())
}
auto_ml_utils.create_auto_ml_job_if_not_exist(sagemaker_session)
auto_ml = AutoML(role=ROLE,
target_attribute_name=TARGET_ATTRIBUTE_NAME,
sagemaker_session=sagemaker_session)
desc = auto_ml.describe_auto_ml_job(job_name=AUTO_ML_JOB_NAME)
assert desc["AutoMLJobName"] == AUTO_ML_JOB_NAME
assert desc["AutoMLJobStatus"] == "Completed"
assert isinstance(desc["BestCandidate"], dict)
assert desc["InputDataConfig"] == expected_default_input_config
assert desc["AutoMLJobConfig"] == EXPECTED_DEFAULT_JOB_CONFIG
assert desc["OutputDataConfig"] == expected_default_output_config
def test_list_candidates(sagemaker_session):
auto_ml_utils.create_auto_ml_job_if_not_exist(sagemaker_session)
auto_ml = AutoML(role=ROLE,
target_attribute_name=TARGET_ATTRIBUTE_NAME,
sagemaker_session=sagemaker_session)
candidates = auto_ml.list_candidates(job_name=AUTO_ML_JOB_NAME)
assert len(candidates) == 3
def test_best_candidate(sagemaker_session):
auto_ml_utils.create_auto_ml_job_if_not_exist(sagemaker_session)
auto_ml = AutoML(role=ROLE,
target_attribute_name=TARGET_ATTRIBUTE_NAME,
sagemaker_session=sagemaker_session)
best_candidate = auto_ml.best_candidate(job_name=AUTO_ML_JOB_NAME)
assert len(best_candidate["InferenceContainers"]) == 3
assert len(best_candidate["CandidateSteps"]) == 4
assert best_candidate["CandidateStatus"] == "Completed"
def test_candidate_estimator_get_steps(sagemaker_session):
auto_ml_utils.create_auto_ml_job_if_not_exist(sagemaker_session)
auto_ml = AutoML(role=ROLE,
target_attribute_name=TARGET_ATTRIBUTE_NAME,
sagemaker_session=sagemaker_session)
candidates = auto_ml.list_candidates(job_name=AUTO_ML_JOB_NAME)
candidate = candidates[1]
candidate_estimator = CandidateEstimator(candidate, sagemaker_session)
steps = candidate_estimator.get_steps()
assert len(steps) == 3
def test_deploy_best_candidate(sagemaker_session, cpu_instance_type):
auto_ml_utils.create_auto_ml_job_if_not_exist(sagemaker_session)
auto_ml = AutoML(
role=ROLE, target_attribute_name=TARGET_ATTRIBUTE_NAME, sagemaker_session=sagemaker_session
)
best_candidate = auto_ml.best_candidate(job_name=AUTO_ML_JOB_NAME)
endpoint_name = unique_name_from_base("sagemaker-auto-ml-best-candidate-test")
with timeout(minutes=AUTO_ML_DEFAULT_TIMEMOUT_MINUTES):
auto_ml.deploy(
candidate=best_candidate,
initial_instance_count=INSTANCE_COUNT,
instance_type=cpu_instance_type,
endpoint_name=endpoint_name,
)
endpoint_status = sagemaker_session.sagemaker_client.describe_endpoint(
EndpointName=endpoint_name
)["EndpointStatus"]
assert endpoint_status == "InService"
sagemaker_session.sagemaker_client.delete_endpoint(EndpointName=endpoint_name)
