Update deploy method

Author: Balaji Veeramani

src/sagemaker/automl/automl.py

@@ -337,6 +337,8 @@ def deploy(
         self,
         initial_instance_count,
         instance_type,
+        serializer=None,
+        deserializer=None,
         candidate=None,
         sagemaker_session=None,
         name=None,
@@ -356,6 +358,14 @@ def deploy(
                 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.
+            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.
             candidate (CandidateEstimator or dict): a CandidateEstimator used for deploying
                 to a SageMaker Inference Pipeline. If None, the best candidate will
                 be used. If the candidate input is a dict, a CandidateEstimator will be
@@ -405,6 +415,8 @@ def deploy(
         return model.deploy(
             initial_instance_count=initial_instance_count,
             instance_type=instance_type,
+            serializer=serializer,
+            deserializer=deserializer,
             endpoint_name=endpoint_name,
             tags=tags,
             wait=wait,

src/sagemaker/estimator.py

@@ -644,6 +644,8 @@ def deploy(
         self,
         initial_instance_count,
         instance_type,
+        serializer=None,
+        deserializer=None,
         accelerator_type=None,
         endpoint_name=None,
         use_compiled_model=False,
@@ -665,6 +667,14 @@ def deploy(
                 deploy to an endpoint for prediction.
             instance_type (str): Type of EC2 instance to deploy to an endpoint
                 for prediction, for example, 'ml.c4.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.
+            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.
             accelerator_type (str): Type of Elastic Inference accelerator to
                 attach to an endpoint for model loading and inference, for
                 example, 'ml.eia1.medium'. If not specified, no Elastic
@@ -727,6 +737,8 @@ def deploy(
         return model.deploy(
             instance_type=instance_type,
             initial_instance_count=initial_instance_count,
+            serializer=serializer,
+            deserializer=deserializer,
             accelerator_type=accelerator_type,
             endpoint_name=endpoint_name,
             tags=tags or self.tags,
@@ -1341,10 +1353,6 @@ def create_model(
         role=None,
         image_uri=None,
         predictor_cls=None,
-        serializer=None,
-        deserializer=None,
-        content_type=None,
-        accept=None,
         vpc_config_override=vpc_utils.VPC_CONFIG_DEFAULT,
         **kwargs
     ):
@@ -1362,17 +1370,6 @@ def create_model(
                 Defaults to the image used for training.
             predictor_cls (Predictor): The predictor class to use when
                 deploying the model.
-            serializer (callable): Should accept a single argument, the input
-                data, and return a sequence of bytes. May provide a content_type
-                attribute that defines the endpoint request content type
-            deserializer (callable): Should accept two arguments, the result
-                data and the response content type, and return a sequence of
-                bytes. May provide a content_type attribute that defines th
-                endpoint response Accept content type.
-            content_type (str): The invocation ContentType, overriding any
-                content_type from the serializer
-            accept (str): The invocation Accept, overriding any accept from the
-                deserializer.
             vpc_config_override (dict[str, list[str]]): Optional override for VpcConfig set on
                 the model.
                 Default: use subnets and security groups from this Estimator.
@@ -1391,7 +1388,7 @@ def create_model(
         if predictor_cls is None:
 
             def predict_wrapper(endpoint, session):
-                return Predictor(endpoint, session, serializer, deserializer, content_type, accept)
+                return Predictor(endpoint, session)
 
             predictor_cls = predict_wrapper
 

src/sagemaker/model.py

@@ -420,6 +420,8 @@ def deploy(
         self,
         initial_instance_count,
         instance_type,
+        serializer=None,
+        deserializer=None,
         accelerator_type=None,
         endpoint_name=None,
         tags=None,
@@ -446,6 +448,14 @@ def deploy(
                 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.
+            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.
             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
@@ -512,7 +522,12 @@ def deploy(
         )
 
         if self.predictor_cls:
-            return self.predictor_cls(self.endpoint_name, self.sagemaker_session)
+            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 transformer(

src/sagemaker/multidatamodel.py

@@ -143,6 +143,8 @@ def deploy(
         self,
         initial_instance_count,
         instance_type,
+        serializer=None,
+        deserializer=None,
         accelerator_type=None,
         endpoint_name=None,
         tags=None,
@@ -171,6 +173,14 @@ def deploy(
                 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.
+            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.
             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
@@ -246,7 +256,12 @@ def deploy(
         )
 
         if predictor:
-            return predictor(self.endpoint_name, self.sagemaker_session)
+            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 add_model(self, model_data_source, model_data_path=None):

src/sagemaker/pipeline.py

@@ -85,6 +85,8 @@ def deploy(
         self,
         initial_instance_count,
         instance_type,
+        serializer=None,
+        deserializer=None,
         endpoint_name=None,
         tags=None,
         wait=True,
@@ -110,6 +112,14 @@ def deploy(
                 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.
+            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.
             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
@@ -171,7 +181,12 @@ def deploy(
             )
 
         if self.predictor_cls:
-            return self.predictor_cls(self.endpoint_name, self.sagemaker_session)
+            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 _create_sagemaker_pipeline_model(self, instance_type):

src/sagemaker/tuner.py

@@ -672,6 +672,8 @@ def deploy(
         self,
         initial_instance_count,
         instance_type,
+        serializer=None,
+        deserializer=None,
         accelerator_type=None,
         endpoint_name=None,
         wait=True,
@@ -691,6 +693,14 @@ def deploy(
                 deploy to an endpoint for prediction.
             instance_type (str): Type of EC2 instance to deploy to an endpoint
                 for prediction, for example, 'ml.c4.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.
+            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.
             accelerator_type (str): Type of Elastic Inference accelerator to
                 attach to an endpoint for model loading and inference, for
                 example, 'ml.eia1.medium'. If not specified, no Elastic
@@ -725,6 +735,8 @@ def deploy(
         return best_estimator.deploy(
             initial_instance_count=initial_instance_count,
             instance_type=instance_type,
+            serializer=serializer,
+            deserializer=deserializer,
             accelerator_type=accelerator_type,
             endpoint_name=endpoint_name or best_training_job["TrainingJobName"],
             wait=wait,