def test_endpoint_from_production_variants(sagemaker_session):
ims = sagemaker_session
ims.sagemaker_client.describe_endpoint = Mock(
return_value={'EndpointStatus': 'InService'})
pvs = [
sagemaker.production_variant('A', 'ml.p2.xlarge'),
sagemaker.production_variant('B', 'p299.4096xlarge')
]
ex = ClientError(
{
'Error': {
'Code': 'ValidationException',
'Message': 'Could not find your thing'
}
}, 'b')
ims.sagemaker_client.describe_endpoint_config = Mock(side_effect=ex)
sagemaker_session.endpoint_from_production_variants('some-endpoint', pvs)
sagemaker_session.sagemaker_client.create_endpoint.assert_called_with(
EndpointConfigName='some-endpoint', EndpointName='some-endpoint')
sagemaker_session.sagemaker_client.create_endpoint_config.assert_called_with(
EndpointConfigName='some-endpoint',
ProductionVariants=[{
'InstanceType': 'ml.p2.xlarge',
'ModelName': 'A',
'InitialVariantWeight': 1,
'InitialInstanceCount': 1,
'VariantName': 'AllTraffic'
}, {
'InstanceType': 'p299.4096xlarge',
'ModelName': 'B',
'InitialVariantWeight': 1,
'InitialInstanceCount': 1,
'VariantName': 'AllTraffic'
}])
def test_endpoint_from_production_variants_with_tags(sagemaker_session):
ims = sagemaker_session
ims.sagemaker_client.describe_endpoint = Mock(return_value={'EndpointStatus': 'InService'})
pvs = [sagemaker.production_variant('A', 'ml.p2.xlarge'), sagemaker.production_variant('B', 'p299.4096xlarge')]
ex = ClientError({'Error': {'Code': 'ValidationException', 'Message': 'Could not find your thing'}}, 'b')
ims.sagemaker_client.describe_endpoint_config = Mock(side_effect=ex)
tags = [{'ModelName': 'TestModel'}]
sagemaker_session.endpoint_from_production_variants('some-endpoint', pvs, tags)
sagemaker_session.sagemaker_client.create_endpoint.assert_called_with(EndpointConfigName='some-endpoint',
EndpointName='some-endpoint')
sagemaker_session.sagemaker_client.create_endpoint_config.assert_called_with(
EndpointConfigName='some-endpoint',
ProductionVariants=[
{
'InstanceType': 'ml.p2.xlarge',
'ModelName': 'A',
'InitialVariantWeight': 1,
'InitialInstanceCount': 1,
'VariantName': 'AllTraffic'
},
{
'InstanceType': 'p299.4096xlarge',
'ModelName': 'B',
'InitialVariantWeight': 1,
'InitialInstanceCount': 1,
'VariantName': 'AllTraffic'}],
Tags=tags)
def test_endpoint_from_production_variants_with_accelerator_type(
sagemaker_session):
ims = sagemaker_session
ims.sagemaker_client.describe_endpoint = Mock(
return_value={"EndpointStatus": "InService"})
pvs = [
sagemaker.production_variant("A",
"ml.p2.xlarge",
accelerator_type=ACCELERATOR_TYPE),
sagemaker.production_variant("B",
"p299.4096xlarge",
accelerator_type=ACCELERATOR_TYPE),
]
ex = ClientError(
{
"Error": {
"Code": "ValidationException",
"Message": "Could not find your thing"
}
}, "b")
ims.sagemaker_client.describe_endpoint_config = Mock(side_effect=ex)
tags = [{"ModelName": "TestModel"}]
sagemaker_session.endpoint_from_production_variants(
"some-endpoint", pvs, tags)
sagemaker_session.sagemaker_client.create_endpoint.assert_called_with(
EndpointConfigName="some-endpoint",
EndpointName="some-endpoint",
Tags=tags)
sagemaker_session.sagemaker_client.create_endpoint_config.assert_called_with(
EndpointConfigName="some-endpoint", ProductionVariants=pvs, Tags=tags)
def test_endpoint_from_production_variants_with_accelerator_type(
sagemaker_session):
ims = sagemaker_session
ims.sagemaker_client.describe_endpoint = Mock(
return_value={'EndpointStatus': 'InService'})
pvs = [
sagemaker.production_variant('A',
'ml.p2.xlarge',
accelerator_type=ACCELERATOR_TYPE),
sagemaker.production_variant('B',
'p299.4096xlarge',
accelerator_type=ACCELERATOR_TYPE)
]
ex = ClientError(
{
'Error': {
'Code': 'ValidationException',
'Message': 'Could not find your thing'
}
}, 'b')
ims.sagemaker_client.describe_endpoint_config = Mock(side_effect=ex)
tags = [{'ModelName': 'TestModel'}]
sagemaker_session.endpoint_from_production_variants(
'some-endpoint', pvs, tags)
sagemaker_session.sagemaker_client.create_endpoint.assert_called_with(
EndpointConfigName='some-endpoint', EndpointName='some-endpoint')
sagemaker_session.sagemaker_client.create_endpoint_config.assert_called_with(
EndpointConfigName='some-endpoint', ProductionVariants=pvs, Tags=tags)
def deploy(self,
initial_instance_count,
instance_type,
endpoint_name=None,
tags=None,
wait=True):
"""Deploy this ``Model`` to an ``Endpoint`` and optionally return a
``Predictor``.
Create a SageMaker ``Model`` and ``EndpointConfig``, and deploy an
``Endpoint`` from this ``Model``. If ``self.predictor_cls`` is not None,
this method returns a the result of invoking ``self.predictor_cls`` on
the created endpoint name.
The name of the created model is accessible in the ``name`` field of
this ``Model`` after deploy returns
The name of the created endpoint is accessible in the
``endpoint_name`` field of this ``Model`` after deploy returns.
Args:
initial_instance_count (int): The initial number of instances to run
in the ``Endpoint`` created from this ``Model``.
instance_type (str): The EC2 instance type to deploy this Model to.
For example, 'ml.p2.xlarge'.
endpoint_name (str): The name of the endpoint to create (default:
None). If not specified, a unique endpoint name will be created.
tags (List[dict[str, str]]): The list of tags to attach to this
specific endpoint.
wait (bool): Whether the call should wait until the deployment of
model completes (default: True).
Returns:
callable[string, sagemaker.session.Session] or None: Invocation of
``self.predictor_cls`` on the created endpoint name, if ``self.predictor_cls``
is not None. Otherwise, return None.
"""
if not self.sagemaker_session:
self.sagemaker_session = Session()
containers = self.pipeline_container_def(instance_type)
self.name = self.name or name_from_image(containers[0]["Image"])
self.sagemaker_session.create_model(self.name,
self.role,
containers,
vpc_config=self.vpc_config)
production_variant = sagemaker.production_variant(
self.name, instance_type, initial_instance_count)
self.endpoint_name = endpoint_name or self.name
self.sagemaker_session.endpoint_from_production_variants(
self.endpoint_name, [production_variant], tags, wait=wait)
if self.predictor_cls:
return self.predictor_cls(self.endpoint_name,
self.sagemaker_session)
return None
def deploy_config(model,
initial_instance_count,
instance_type,
endpoint_name=None,
tags=None):
"""Export Airflow deploy config from a SageMaker model
Args:
model (sagemaker.model.Model): The SageMaker model to export the Airflow
config from.
initial_instance_count (int): The initial number of instances to run in
the ``Endpoint`` created from this ``Model``.
instance_type (str): The EC2 instance type to deploy this Model to. For
example, 'ml.p2.xlarge'.
endpoint_name (str): The name of the endpoint to create (default: None).
If not specified, a unique endpoint name will be created.
tags (list[dict]): List of tags for labeling a training job. For more,
see https://docs.aws.amazon.com/sagemaker/latest/dg/API_Tag.html.
Returns:
dict: Deploy config that can be directly used by
SageMakerEndpointOperator in Airflow.
"""
model_base_config = model_config(instance_type, model)
production_variant = sagemaker.production_variant(model.name,
instance_type,
initial_instance_count)
name = model.name
config_options = {
"EndpointConfigName": name,
"ProductionVariants": [production_variant]
}
if tags is not None:
config_options["Tags"] = tags
endpoint_name = endpoint_name or name
endpoint_base_config = {
"EndpointName": endpoint_name,
"EndpointConfigName": name
}
config = {
"Model": model_base_config,
"EndpointConfig": config_options,
"Endpoint": endpoint_base_config,
}
# if there is s3 operations needed for model, move it to root level of config
s3_operations = model_base_config.pop("S3Operations", None)
if s3_operations is not None:
config["S3Operations"] = s3_operations
return config
def deploy(self, initial_instance_count, instance_type, accelerator_type=None, endpoint_name=None, tags=None):
"""Deploy this ``Model`` to an ``Endpoint`` and optionally return a ``Predictor``.
Create a SageMaker ``Model`` and ``EndpointConfig``, and deploy an ``Endpoint`` from this ``Model``.
If ``self.predictor_cls`` is not None, this method returns a the result of invoking
``self.predictor_cls`` on the created endpoint name.
The name of the created model is accessible in the ``name`` field of this ``Model`` after deploy returns
The name of the created endpoint is accessible in the ``endpoint_name``
field of this ``Model`` after deploy returns.
Args:
instance_type (str): The EC2 instance type to deploy this Model to. For example, 'ml.p2.xlarge'.
initial_instance_count (int): The initial number of instances to run in the
``Endpoint`` created from this ``Model``.
accelerator_type (str): Type of Elastic Inference accelerator to deploy this model for model loading
and inference, for example, 'ml.eia1.medium'. If not specified, no Elastic Inference accelerator
will be attached to the endpoint.
For more information: https://docs.aws.amazon.com/sagemaker/latest/dg/ei.html
endpoint_name (str): The name of the endpoint to create (default: None).
If not specified, a unique endpoint name will be created.
tags(List[dict[str, str]]): The list of tags to attach to this specific endpoint.
Returns:
callable[string, sagemaker.session.Session] or None: Invocation of ``self.predictor_cls`` on
the created endpoint name, if ``self.predictor_cls`` is not None. Otherwise, return None.
"""
if not self.sagemaker_session:
if instance_type in ('local', 'local_gpu'):
self.sagemaker_session = local.LocalSession()
else:
self.sagemaker_session = session.Session()
if self.role is None:
raise ValueError("Role can not be null for deploying a model")
compiled_model_suffix = '-'.join(instance_type.split('.')[:-1])
if self._is_compiled_model:
self.name += compiled_model_suffix
self._create_sagemaker_model(instance_type, accelerator_type)
production_variant = sagemaker.production_variant(self.name, instance_type, initial_instance_count,
accelerator_type=accelerator_type)
if endpoint_name:
self.endpoint_name = endpoint_name
else:
self.endpoint_name = self.name
if self._is_compiled_model and not self.endpoint_name.endswith(compiled_model_suffix):
self.endpoint_name += compiled_model_suffix
self.sagemaker_session.endpoint_from_production_variants(self.endpoint_name, [production_variant], tags)
if self.predictor_cls:
return self.predictor_cls(self.endpoint_name, self.sagemaker_session)
def deploy(self,
initial_instance_count,
instance_type,
endpoint_name=None,
tags=None):
"""Deploy this ``Model`` to an ``Endpoint`` and optionally return a ``Predictor``.
Create a SageMaker ``Model`` and ``EndpointConfig``, and deploy an ``Endpoint`` from this ``Model``.
If ``self.predictor_cls`` is not None, this method returns a the result of invoking
``self.predictor_cls`` on the created endpoint name.
The name of the created model is accessible in the ``name`` field of this ``Model`` after deploy returns
The name of the created endpoint is accessible in the ``endpoint_name``
field of this ``Model`` after deploy returns.
Args:
instance_type (str): The EC2 instance type to deploy this Model to. For example, 'ml.p2.xlarge'.
initial_instance_count (int): The initial number of instances to run in the
``Endpoint`` created from this ``Model``.
endpoint_name (str): The name of the endpoint to create (default: None).
If not specified, a unique endpoint name will be created.
Returns:
callable[string, sagemaker.session.Session] or None: Invocation of ``self.predictor_cls`` on
the created endpoint name, if ``self.predictor_cls`` is not None. Otherwise, return None.
"""
if not self.sagemaker_session:
if instance_type in ('local', 'local_gpu'):
self.sagemaker_session = LocalSession()
else:
self.sagemaker_session = Session()
container_def = self.prepare_container_def(instance_type)
self.name = self.name or name_from_image(container_def['Image'])
self.sagemaker_session.create_model(self.name,
self.role,
container_def,
vpc_config=self.vpc_config)
production_variant = sagemaker.production_variant(
self.name, instance_type, initial_instance_count)
self.endpoint_name = endpoint_name or self.name
self.sagemaker_session.endpoint_from_production_variants(
self.endpoint_name, [production_variant], tags)
if self.predictor_cls:
return self.predictor_cls(self.endpoint_name,
self.sagemaker_session)
def deploy(self, initial_instance_count, instance_type, endpoint_name=None, tags=None):
"""Deploy this ``Model`` to an ``Endpoint`` and optionally return a ``Predictor``.
Create a SageMaker ``Model`` and ``EndpointConfig``, and deploy an ``Endpoint`` from this ``Model``.
If ``self.predictor_cls`` is not None, this method returns a the result of invoking
``self.predictor_cls`` on the created endpoint name.
The name of the created endpoint is accessible in the ``endpoint_name``
field of this ``Model`` after deploy returns.
Args:
instance_type (str): The EC2 instance type to deploy this Model to. For example, 'ml.p2.xlarge'.
initial_instance_count (int): The initial number of instances to run in the
``Endpoint`` created from this ``Model``.
endpoint_name (str): The name of the endpoint to create (default: None).
If not specified, a unique endpoint name will be created.
tags (list[dict[str, str]]): A list of key-value pairs for tagging the endpoint (default: None).
Returns:
callable[string, sagemaker.session.Session] or None: Invocation of ``self.predictor_cls`` on
the created endpoint name, if ``self.predictor_cls`` is not None. Otherwise, return None.
"""
if not self.sagemaker_session:
if instance_type in ('local', 'local_gpu'):
self.sagemaker_session = LocalSession()
else:
self.sagemaker_session = Session()
container_def = self.prepare_container_def(instance_type)
model_name = self.name or name_from_image(container_def['Image'])
self.sagemaker_session.create_model(model_name, self.role, container_def)
production_variant = sagemaker.production_variant(model_name, instance_type, initial_instance_count)
self.endpoint_name = endpoint_name or model_name
self.sagemaker_session.endpoint_from_production_variants(self.endpoint_name, [production_variant], tags)
if self.predictor_cls:
return self.predictor_cls(self.endpoint_name, self.sagemaker_session)
def deploy(
self,
initial_instance_count,
instance_type,
accelerator_type=None,
endpoint_name=None,
update_endpoint=False,
tags=None,
kms_key=None,
wait=True,
):
"""Deploy this ``Model`` to an ``Endpoint`` and optionally return a ``Predictor``.
Create a SageMaker ``Model`` and ``EndpointConfig``, and deploy an ``Endpoint`` from this ``Model``.
If ``self.predictor_cls`` is not None, this method returns a the result of invoking
``self.predictor_cls`` on the created endpoint name.
The name of the created model is accessible in the ``name`` field of this ``Model`` after deploy returns
The name of the created endpoint is accessible in the ``endpoint_name``
field of this ``Model`` after deploy returns.
Args:
instance_type (str): The EC2 instance type to deploy this Model to. For example, 'ml.p2.xlarge'.
initial_instance_count (int): The initial number of instances to run in the
``Endpoint`` created from this ``Model``.
accelerator_type (str): Type of Elastic Inference accelerator to deploy this model for model loading
and inference, for example, 'ml.eia1.medium'. If not specified, no Elastic Inference accelerator
will be attached to the endpoint.
For more information: https://docs.aws.amazon.com/sagemaker/latest/dg/ei.html
endpoint_name (str): The name of the endpoint to create (default: None).
If not specified, a unique endpoint name will be created.
update_endpoint (bool): Flag to update the model in an existing Amazon SageMaker endpoint.
If True, this will deploy a new EndpointConfig to an already existing endpoint and delete resources
corresponding to the previous EndpointConfig. If False, a new endpoint will be created. Default: False
tags(List[dict[str, str]]): The list of tags to attach to this specific endpoint.
kms_key (str): The ARN of the KMS key that is used to encrypt the data on the
storage volume attached to the instance hosting the endpoint.
wait (bool): Whether the call should wait until the deployment of this model completes (default: True).
Returns:
callable[string, sagemaker.session.Session] or None: Invocation of ``self.predictor_cls`` on
the created endpoint name, if ``self.predictor_cls`` is not None. Otherwise, return None.
"""
if not self.sagemaker_session:
if instance_type in ("local", "local_gpu"):
self.sagemaker_session = local.LocalSession()
else:
self.sagemaker_session = session.Session()
if self.role is None:
raise ValueError("Role can not be null for deploying a model")
compiled_model_suffix = "-".join(instance_type.split(".")[:-1])
if self._is_compiled_model:
self.name += compiled_model_suffix
self._create_sagemaker_model(instance_type, accelerator_type, tags)
production_variant = sagemaker.production_variant(
self.name,
instance_type,
initial_instance_count,
accelerator_type=accelerator_type)
if endpoint_name:
self.endpoint_name = endpoint_name
else:
self.endpoint_name = self.name
if self._is_compiled_model and not self.endpoint_name.endswith(
compiled_model_suffix):
self.endpoint_name += compiled_model_suffix
if update_endpoint:
endpoint_config_name = self.sagemaker_session.create_endpoint_config(
name=self.name,
model_name=self.name,
initial_instance_count=initial_instance_count,
instance_type=instance_type,
accelerator_type=accelerator_type,
tags=tags,
kms_key=kms_key,
)
self.sagemaker_session.update_endpoint(self.endpoint_name,
endpoint_config_name)
else:
self.sagemaker_session.endpoint_from_production_variants(
self.endpoint_name, [production_variant], tags, kms_key, wait)
if self.predictor_cls:
return self.predictor_cls(self.endpoint_name,
self.sagemaker_session)
def deploy(
self,
initial_instance_count,
instance_type,
serializer=None,
deserializer=None,
endpoint_name=None,
tags=None,
wait=True,
update_endpoint=False,
data_capture_config=None,
):
"""Deploy this ``Model`` to an ``Endpoint`` and optionally return a
``Predictor``.
Create a SageMaker ``Model`` and ``EndpointConfig``, and deploy an
``Endpoint`` from this ``Model``. If ``self.predictor_cls`` is not None,
this method returns a the result of invoking ``self.predictor_cls`` on
the created endpoint name.
The name of the created model is accessible in the ``name`` field of
this ``Model`` after deploy returns
The name of the created endpoint is accessible in the
``endpoint_name`` field of this ``Model`` after deploy returns.
Args:
initial_instance_count (int): The initial number of instances to run
in the ``Endpoint`` created from this ``Model``.
instance_type (str): The EC2 instance type to deploy this Model to.
For example, 'ml.p2.xlarge'.
serializer (:class:`~sagemaker.serializers.BaseSerializer`): A
serializer object, used to encode data for an inference endpoint
(default: None). If ``serializer`` is not None, then
``serializer`` will override the default serializer. The
default serializer is set by the ``predictor_cls``.
deserializer (:class:`~sagemaker.deserializers.BaseDeserializer`): A
deserializer object, used to decode data from an inference
endpoint (default: None). If ``deserializer`` is not None, then
``deserializer`` will override the default deserializer. The
default deserializer is set by the ``predictor_cls``.
endpoint_name (str): The name of the endpoint to create (default:
None). If not specified, a unique endpoint name will be created.
tags (List[dict[str, str]]): The list of tags to attach to this
specific endpoint.
wait (bool): Whether the call should wait until the deployment of
model completes (default: True).
update_endpoint (bool): Flag to update the model in an existing
Amazon SageMaker endpoint. If True, this will deploy a new
EndpointConfig to an already existing endpoint and delete
resources corresponding to the previous EndpointConfig. If
False, a new endpoint will be created. Default: False
data_capture_config (sagemaker.model_monitor.DataCaptureConfig): Specifies
configuration related to Endpoint data capture for use with
Amazon SageMaker Model Monitoring. Default: None.
Returns:
callable[string, sagemaker.session.Session] or None: Invocation of
``self.predictor_cls`` on the created endpoint name, if ``self.predictor_cls``
is not None. Otherwise, return None.
"""
if not self.sagemaker_session:
self.sagemaker_session = Session()
containers = self.pipeline_container_def(instance_type)
self.name = self.name or name_from_image(containers[0]["Image"])
self.sagemaker_session.create_model(self.name,
self.role,
containers,
vpc_config=self.vpc_config)
production_variant = sagemaker.production_variant(
self.name, instance_type, initial_instance_count)
self.endpoint_name = endpoint_name or self.name
data_capture_config_dict = None
if data_capture_config is not None:
data_capture_config_dict = data_capture_config._to_request_dict()
if update_endpoint:
endpoint_config_name = self.sagemaker_session.create_endpoint_config(
name=self.name,
model_name=self.name,
initial_instance_count=initial_instance_count,
instance_type=instance_type,
tags=tags,
data_capture_config_dict=data_capture_config_dict,
)
self.sagemaker_session.update_endpoint(self.endpoint_name,
endpoint_config_name,
wait=wait)
else:
self.sagemaker_session.endpoint_from_production_variants(
name=self.endpoint_name,
production_variants=[production_variant],
tags=tags,
wait=wait,
data_capture_config_dict=data_capture_config_dict,
)
if self.predictor_cls:
predictor = self.predictor_cls(self.endpoint_name,
self.sagemaker_session)
if serializer:
predictor.serializer = serializer
if deserializer:
predictor.deserializer = deserializer
return predictor
return None
def deploy(
self,
initial_instance_count,
instance_type,
accelerator_type=None,
endpoint_name=None,
update_endpoint=False,
tags=None,
kms_key=None,
wait=True,
data_capture_config=None,
):
"""Deploy this ``Model`` to an ``Endpoint`` and optionally return a
``Predictor``.
Create a SageMaker ``Model`` and ``EndpointConfig``, and deploy an
``Endpoint`` from this ``Model``. If ``self.predictor_cls`` is not None,
this method returns a the result of invoking ``self.predictor_cls`` on
the created endpoint name.
The name of the created model is accessible in the ``name`` field of
this ``Model`` after deploy returns
The name of the created endpoint is accessible in the
``endpoint_name`` field of this ``Model`` after deploy returns.
Args:
initial_instance_count (int): The initial number of instances to run
in the ``Endpoint`` created from this ``Model``.
instance_type (str): The EC2 instance type to deploy this Model to.
For example, 'ml.p2.xlarge', or 'local' for local mode.
accelerator_type (str): Type of Elastic Inference accelerator to
deploy this model for model loading and inference, for example,
'ml.eia1.medium'. If not specified, no Elastic Inference
accelerator will be attached to the endpoint. For more
information:
https://docs.aws.amazon.com/sagemaker/latest/dg/ei.html
endpoint_name (str): The name of the endpoint to create (default:
None). If not specified, a unique endpoint name will be created.
update_endpoint (bool): Flag to update the model in an existing
Amazon SageMaker endpoint. If True, this will deploy a new
EndpointConfig to an already existing endpoint and delete
resources corresponding to the previous EndpointConfig. If
False, a new endpoint will be created. Default: False
tags (List[dict[str, str]]): The list of tags to attach to this
specific endpoint.
kms_key (str): The ARN of the KMS key that is used to encrypt the
data on the storage volume attached to the instance hosting the
endpoint.
wait (bool): Whether the call should wait until the deployment of
this model completes (default: True).
data_capture_config (sagemaker.model_monitor.DataCaptureConfig): Specifies
configuration related to Endpoint data capture for use with
Amazon SageMaker Model Monitoring. Default: None.
Returns:
callable[string, sagemaker.session.Session] or None: Invocation of
``self.predictor_cls`` on the created endpoint name, if ``self.predictor_cls``
is not None. Otherwise, return None.
"""
self._init_sagemaker_session_if_does_not_exist(instance_type)
if self.role is None:
raise ValueError("Role can not be null for deploying a model")
if instance_type.startswith("ml.inf") and not self._is_compiled_model:
LOGGER.warning(
"Your model is not compiled. Please compile your model before using Inferentia."
)
compiled_model_suffix = "-".join(instance_type.split(".")[:-1])
if self._is_compiled_model:
name_prefix = self.name or utils.name_from_image(self.image)
self.name = "{}{}".format(name_prefix, compiled_model_suffix)
self._create_sagemaker_model(instance_type, accelerator_type, tags)
production_variant = sagemaker.production_variant(
self.name, instance_type, initial_instance_count, accelerator_type=accelerator_type
)
if endpoint_name:
self.endpoint_name = endpoint_name
else:
self.endpoint_name = self.name
if self._is_compiled_model and not self.endpoint_name.endswith(compiled_model_suffix):
self.endpoint_name += compiled_model_suffix
data_capture_config_dict = None
if data_capture_config is not None:
data_capture_config_dict = data_capture_config._to_request_dict()
if update_endpoint:
endpoint_config_name = self.sagemaker_session.create_endpoint_config(
name=self.name,
model_name=self.name,
initial_instance_count=initial_instance_count,
instance_type=instance_type,
accelerator_type=accelerator_type,
tags=tags,
kms_key=kms_key,
data_capture_config_dict=data_capture_config_dict,
)
self.sagemaker_session.update_endpoint(
self.endpoint_name, endpoint_config_name, wait=wait
)
else:
self.sagemaker_session.endpoint_from_production_variants(
name=self.endpoint_name,
production_variants=[production_variant],
tags=tags,
kms_key=kms_key,
wait=wait,
data_capture_config_dict=data_capture_config_dict,
)
if self.predictor_cls:
return self.predictor_cls(self.endpoint_name, self.sagemaker_session)
return None
def deploy(
self,
initial_instance_count,
instance_type,
serializer=None,
deserializer=None,
accelerator_type=None,
endpoint_name=None,
tags=None,
kms_key=None,
wait=True,
data_capture_config=None,
):
"""Deploy this ``Model`` to an ``Endpoint`` and optionally return a ``Predictor``.
Create a SageMaker ``Model`` and ``EndpointConfig``, and deploy an
``Endpoint`` from this ``Model``. If self.model is not None, then the ``Endpoint``
will be deployed with parameters in self.model (like vpc_config,
enable_network_isolation, etc). If self.model is None, then use the parameters
in ``MultiDataModel`` constructor will be used. If ``self.predictor_cls`` is not
None, this method returns a the result of invoking ``self.predictor_cls`` on
the created endpoint name.
The name of the created model is accessible in the ``name`` field of
this ``Model`` after deploy returns
The name of the created endpoint is accessible in the
``endpoint_name`` field of this ``Model`` after deploy returns.
Args:
initial_instance_count (int): The initial number of instances to run
in the ``Endpoint`` created from this ``Model``.
instance_type (str): The EC2 instance type to deploy this Model to.
For example, 'ml.p2.xlarge', or 'local' for local mode.
serializer (:class:`~sagemaker.serializers.BaseSerializer`): A
serializer object, used to encode data for an inference endpoint
(default: None). If ``serializer`` is not None, then
``serializer`` will override the default serializer. The
default serializer is set by the ``predictor_cls``.
deserializer (:class:`~sagemaker.deserializers.BaseDeserializer`): A
deserializer object, used to decode data from an inference
endpoint (default: None). If ``deserializer`` is not None, then
``deserializer`` will override the default deserializer. The
default deserializer is set by the ``predictor_cls``.
accelerator_type (str): Type of Elastic Inference accelerator to
deploy this model for model loading and inference, for example,
'ml.eia1.medium'. If not specified, no Elastic Inference
accelerator will be attached to the endpoint. For more
information:
https://docs.aws.amazon.com/sagemaker/latest/dg/ei.html
endpoint_name (str): The name of the endpoint to create (default:
None). If not specified, a unique endpoint name will be created.
tags (List[dict[str, str]]): The list of tags to attach to this
specific endpoint.
kms_key (str): The ARN of the KMS key that is used to encrypt the
data on the storage volume attached to the instance hosting the
endpoint.
wait (bool): Whether the call should wait until the deployment of
this model completes (default: True).
data_capture_config (sagemaker.model_monitor.DataCaptureConfig): Specifies
configuration related to Endpoint data capture for use with
Amazon SageMaker Model Monitoring. Default: None.
Returns:
callable[string, sagemaker.session.Session] or None: Invocation of
``self.predictor_cls`` on the created endpoint name,
if ``self.predictor_cls``
is not None. Otherwise, return None.
"""
# Set model specific parameters
if self.model:
enable_network_isolation = self.model.enable_network_isolation()
role = self.model.role
vpc_config = self.model.vpc_config
predictor_cls = self.model.predictor_cls
else:
enable_network_isolation = self.enable_network_isolation()
role = self.role
vpc_config = self.vpc_config
predictor_cls = self.predictor_cls
if role is None:
raise ValueError("Role can not be null for deploying a model")
if instance_type == "local" and not isinstance(self.sagemaker_session,
local.LocalSession):
self.sagemaker_session = local.LocalSession()
container_def = self.prepare_container_def(
instance_type, accelerator_type=accelerator_type)
self.sagemaker_session.create_model(
self.name,
role,
container_def,
vpc_config=vpc_config,
enable_network_isolation=enable_network_isolation,
tags=tags,
)
production_variant = sagemaker.production_variant(
self.name,
instance_type,
initial_instance_count,
accelerator_type=accelerator_type)
if endpoint_name:
self.endpoint_name = endpoint_name
else:
self.endpoint_name = self.name
data_capture_config_dict = None
if data_capture_config is not None:
data_capture_config_dict = data_capture_config._to_request_dict()
self.sagemaker_session.endpoint_from_production_variants(
name=self.endpoint_name,
production_variants=[production_variant],
tags=tags,
kms_key=kms_key,
wait=wait,
data_capture_config_dict=data_capture_config_dict,
)
if predictor_cls:
predictor = predictor_cls(self.endpoint_name,
self.sagemaker_session)
if serializer:
predictor.serializer = serializer
if deserializer:
predictor.deserializer = deserializer
return predictor
return None
def deploy(
self,
initial_instance_count,
instance_type,
serializer=None,
deserializer=None,
accelerator_type=None,
endpoint_name=None,
tags=None,
kms_key=None,
wait=True,
data_capture_config=None,
**kwargs,
):
"""Deploy this ``Model`` to an ``Endpoint`` and optionally return a
``Predictor``.
Create a SageMaker ``Model`` and ``EndpointConfig``, and deploy an
``Endpoint`` from this ``Model``. If ``self.predictor_cls`` is not None,
this method returns a the result of invoking ``self.predictor_cls`` on
the created endpoint name.
The name of the created model is accessible in the ``name`` field of
this ``Model`` after deploy returns
The name of the created endpoint is accessible in the
``endpoint_name`` field of this ``Model`` after deploy returns.
Args:
initial_instance_count (int): The initial number of instances to run
in the ``Endpoint`` created from this ``Model``.
instance_type (str): The EC2 instance type to deploy this Model to.
For example, 'ml.p2.xlarge', or 'local' for local mode.
serializer (:class:`~sagemaker.serializers.BaseSerializer`): A
serializer object, used to encode data for an inference endpoint
(default: None). If ``serializer`` is not None, then
``serializer`` will override the default serializer. The
default serializer is set by the ``predictor_cls``.
deserializer (:class:`~sagemaker.deserializers.BaseDeserializer`): A
deserializer object, used to decode data from an inference
endpoint (default: None). If ``deserializer`` is not None, then
``deserializer`` will override the default deserializer. The
default deserializer is set by the ``predictor_cls``.
accelerator_type (str): Type of Elastic Inference accelerator to
deploy this model for model loading and inference, for example,
'ml.eia1.medium'. If not specified, no Elastic Inference
accelerator will be attached to the endpoint. For more
information:
https://docs.aws.amazon.com/sagemaker/latest/dg/ei.html
endpoint_name (str): The name of the endpoint to create (default:
None). If not specified, a unique endpoint name will be created.
tags (List[dict[str, str]]): The list of tags to attach to this
specific endpoint.
kms_key (str): The ARN of the KMS key that is used to encrypt the
data on the storage volume attached to the instance hosting the
endpoint.
wait (bool): Whether the call should wait until the deployment of
this model completes (default: True).
data_capture_config (sagemaker.model_monitor.DataCaptureConfig): Specifies
configuration related to Endpoint data capture for use with
Amazon SageMaker Model Monitoring. Default: None.
Returns:
callable[string, sagemaker.session.Session] or None: Invocation of
``self.predictor_cls`` on the created endpoint name, if ``self.predictor_cls``
is not None. Otherwise, return None.
"""
removed_kwargs("update_endpoint", kwargs)
self._init_sagemaker_session_if_does_not_exist(instance_type)
if self.role is None:
raise ValueError("Role can not be null for deploying a model")
if instance_type.startswith("ml.inf") and not self._is_compiled_model:
LOGGER.warning(
"Your model is not compiled. Please compile your model before using Inferentia."
)
compiled_model_suffix = "-".join(instance_type.split(".")[:-1])
if self._is_compiled_model:
self._ensure_base_name_if_needed(self.image_uri)
if self._base_name is not None:
self._base_name = "-".join(
(self._base_name, compiled_model_suffix))
self._create_sagemaker_model(instance_type, accelerator_type, tags)
production_variant = sagemaker.production_variant(
self.name,
instance_type,
initial_instance_count,
accelerator_type=accelerator_type)
if endpoint_name:
self.endpoint_name = endpoint_name
else:
base_endpoint_name = self._base_name or utils.base_from_name(
self.name)
if self._is_compiled_model and not base_endpoint_name.endswith(
compiled_model_suffix):
base_endpoint_name = "-".join(
(base_endpoint_name, compiled_model_suffix))
self.endpoint_name = utils.name_from_base(base_endpoint_name)
data_capture_config_dict = None
if data_capture_config is not None:
data_capture_config_dict = data_capture_config._to_request_dict()
self.sagemaker_session.endpoint_from_production_variants(
name=self.endpoint_name,
production_variants=[production_variant],
tags=tags,
kms_key=kms_key,
wait=wait,
data_capture_config_dict=data_capture_config_dict,
)
if self.predictor_cls:
predictor = self.predictor_cls(self.endpoint_name,
self.sagemaker_session)
if serializer:
predictor.serializer = serializer
if deserializer:
predictor.deserializer = deserializer
return predictor
return None
