def deploy_model_with_explanations_mock():
with mock.patch.object(
endpoint_service_client_v1beta1.EndpointServiceClient,
"deploy_model") as deploy_model_mock:
deployed_model = gca_endpoint_v1beta1.DeployedModel(
model=_TEST_MODEL_NAME,
display_name=_TEST_DISPLAY_NAME,
)
deploy_model_lro_mock = mock.Mock(ga_operation.Operation)
deploy_model_lro_mock.result.return_value = gca_endpoint_service_v1beta1.DeployModelResponse(
deployed_model=deployed_model, )
deploy_model_mock.return_value = deploy_model_lro_mock
yield deploy_model_mock
def test_deploy_with_explanations(self,
deploy_model_with_explanations_mock,
sync):
aiplatform.init(project=_TEST_PROJECT, location=_TEST_LOCATION)
test_endpoint = models.Endpoint(_TEST_ENDPOINT_NAME)
test_model = models.Model(_TEST_ID)
test_endpoint.deploy(
model=test_model,
machine_type=_TEST_MACHINE_TYPE,
accelerator_type=_TEST_ACCELERATOR_TYPE,
accelerator_count=_TEST_ACCELERATOR_COUNT,
explanation_metadata=_TEST_EXPLANATION_METADATA,
explanation_parameters=_TEST_EXPLANATION_PARAMETERS,
sync=sync,
)
if not sync:
test_endpoint.wait()
expected_machine_spec = gca_machine_resources_v1beta1.MachineSpec(
machine_type=_TEST_MACHINE_TYPE,
accelerator_type=_TEST_ACCELERATOR_TYPE,
accelerator_count=_TEST_ACCELERATOR_COUNT,
)
expected_dedicated_resources = gca_machine_resources_v1beta1.DedicatedResources(
machine_spec=expected_machine_spec,
min_replica_count=1,
max_replica_count=1,
)
expected_deployed_model = gca_endpoint_v1beta1.DeployedModel(
dedicated_resources=expected_dedicated_resources,
model=test_model.resource_name,
display_name=None,
explanation_spec=gca_endpoint_v1beta1.explanation.ExplanationSpec(
metadata=_TEST_EXPLANATION_METADATA,
parameters=_TEST_EXPLANATION_PARAMETERS,
),
)
deploy_model_with_explanations_mock.assert_called_once_with(
endpoint=test_endpoint.resource_name,
deployed_model=expected_deployed_model,
traffic_split={"0": 100},
metadata=(),
)