Merge branch 'json-update-script' of github.com:ChuyangDeng/sagemaker-python-sdk into json-update-script

Author: Chuyang Deng

doc/v2.rst

@@ -94,6 +94,18 @@ Please instantiate the objects instead.
 The ``update_endpoint`` argument in ``deploy()`` methods for estimators and models has been deprecated.
 Please use :func:`sagemaker.predictor.Predictor.update_endpoint` instead.
 
+``content_type`` and ``accept`` in the Predictor Constructor
+------------------------------------------------------------
+
+The ``content_type`` and ``accept`` parameters have been removed from the
+following classes and methods:
+- ``sagemaker.predictor.Predictor``
+- ``sagemaker.estimator.Estimator.create_model``
+- ``sagemaker.algorithms.AlgorithmEstimator.create_model``
+- ``sagemaker.tensorflow.model.TensorFlowPredictor``
+
+Please specify content types in a serializer or deserializer class instead.
+
 ``sagemaker.content_types``
 ---------------------------
 
@@ -115,7 +127,6 @@ write MIME types as a string,
 | ``CONTENT_TYPE_NPY``          | ``"application/x-npy"``        |
 +-------------------------------+--------------------------------+
 
-
 Require ``framework_version`` and ``py_version`` for Frameworks
 ===============================================================
 

src/sagemaker/algorithm.py

@@ -16,7 +16,9 @@
 import sagemaker
 import sagemaker.parameter
 from sagemaker import vpc_utils
+from sagemaker.deserializers import BytesDeserializer
 from sagemaker.estimator import EstimatorBase
+from sagemaker.serializers import IdentitySerializer
 from sagemaker.transformer import Transformer
 from sagemaker.predictor import Predictor
 
@@ -251,37 +253,29 @@ def create_model(
         self,
         role=None,
         predictor_cls=None,
-        serializer=None,
-        deserializer=None,
-        content_type=None,
-        accept=None,
+        serializer=IdentitySerializer(),
+        deserializer=BytesDeserializer(),
         vpc_config_override=vpc_utils.VPC_CONFIG_DEFAULT,
         **kwargs
     ):
         """Create a model to deploy.
 
-        The serializer, deserializer, content_type, and accept arguments are
-        only used to define a default Predictor. They are ignored if an
-        explicit predictor class is passed in. Other arguments are passed
-        through to the Model class.
+        The serializer and deserializer are only used to define a default
+        Predictor. They are ignored if an explicit predictor class is passed in.
+        Other arguments are passed through to the Model class.
 
         Args:
             role (str): The ``ExecutionRoleArn`` IAM Role ARN for the ``Model``,
                 which is also used during transform jobs. If not specified, the
                 role from the Estimator will be used.
             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 the
-                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.
+            serializer (:class:`~sagemaker.serializers.BaseSerializer`): A
+                serializer object, used to encode data for an inference endpoint
+                (default: :class:`~sagemaker.serializers.IdentitySerializer`).
+            deserializer (:class:`~sagemaker.deserializers.BaseDeserializer`): A
+                deserializer object, used to decode data from an inference
+                endpoint (default: :class:`~sagemaker.deserializers.BytesDeserializer`).
             vpc_config_override (dict[str, list[str]]): Optional override for VpcConfig set on
                 the model. Default: use subnets and security groups from this Estimator.
                 * 'Subnets' (list[str]): List of subnet ids.
@@ -300,7 +294,7 @@ 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, serializer, deserializer)
 
             predictor_cls = predict_wrapper
 

src/sagemaker/chainer/estimator.py

@@ -32,7 +32,7 @@
 class Chainer(Framework):
     """Handle end-to-end training and deployment of custom Chainer code."""
 
-    __framework_name__ = "chainer"
+    _framework_name = "chainer"
 
     # Hyperparameters
     _use_mpi = "sagemaker_use_mpi"
@@ -131,7 +131,7 @@ def __init__(
         validate_version_or_image_args(framework_version, py_version, image_uri)
         if py_version == "py2":
             logger.warning(
-                python_deprecation_warning(self.__framework_name__, defaults.LATEST_PY2_VERSION)
+                python_deprecation_warning(self._framework_name, defaults.LATEST_PY2_VERSION)
             )
         self.framework_version = framework_version
         self.py_version = py_version
@@ -272,7 +272,7 @@ class constructor
             init_params["image_uri"] = image_uri
             return init_params
 
-        if framework != cls.__framework_name__:
+        if framework != cls._framework_name:
             raise ValueError(
                 "Training job: {} didn't use image for requested framework".format(
                     job_details["TrainingJobName"]

src/sagemaker/chainer/model.py

@@ -59,7 +59,7 @@ class ChainerModel(FrameworkModel):
     ``Endpoint``.
     """
 
-    __framework_name__ = "chainer"
+    _framework_name = "chainer"
 
     def __init__(
         self,
@@ -116,7 +116,7 @@ def __init__(
         validate_version_or_image_args(framework_version, py_version, image_uri)
         if py_version == "py2":
             logger.warning(
-                python_deprecation_warning(self.__framework_name__, defaults.LATEST_PY2_VERSION)
+                python_deprecation_warning(self._framework_name, defaults.LATEST_PY2_VERSION)
             )
         self.framework_version = framework_version
         self.py_version = py_version
@@ -176,7 +176,7 @@ def serving_image_uri(self, region_name, instance_type, accelerator_type=None):
 
         """
         return image_uris.retrieve(
-            self.__framework_name__,
+            self._framework_name,
             region_name,
             version=self.framework_version,
             py_version=self.py_version,

src/sagemaker/debugger.py

@@ -23,32 +23,9 @@
 
 import smdebug_rulesconfig as rule_configs  # noqa: F401 # pylint: disable=unused-import
 
-from sagemaker.utils import get_ecr_image_uri_prefix
-
-RULES_ECR_REPO_NAME = "sagemaker-debugger-rules"
-
-SAGEMAKER_RULE_CONTAINERS_ACCOUNTS_MAP = {
-    "eu-north-1": {RULES_ECR_REPO_NAME: "314864569078"},
-    "me-south-1": {RULES_ECR_REPO_NAME: "986000313247"},
-    "ap-south-1": {RULES_ECR_REPO_NAME: "904829902805"},
-    "eu-west-3": {RULES_ECR_REPO_NAME: "447278800020"},
-    "us-east-2": {RULES_ECR_REPO_NAME: "915447279597"},
-    "eu-west-1": {RULES_ECR_REPO_NAME: "929884845733"},
-    "eu-central-1": {RULES_ECR_REPO_NAME: "482524230118"},
-    "sa-east-1": {RULES_ECR_REPO_NAME: "818342061345"},
-    "ap-east-1": {RULES_ECR_REPO_NAME: "199566480951"},
-    "us-east-1": {RULES_ECR_REPO_NAME: "503895931360"},
-    "ap-northeast-2": {RULES_ECR_REPO_NAME: "578805364391"},
-    "eu-west-2": {RULES_ECR_REPO_NAME: "250201462417"},
-    "ap-northeast-1": {RULES_ECR_REPO_NAME: "430734990657"},
-    "us-west-2": {RULES_ECR_REPO_NAME: "895741380848"},
-    "us-west-1": {RULES_ECR_REPO_NAME: "685455198987"},
-    "ap-southeast-1": {RULES_ECR_REPO_NAME: "972752614525"},
-    "ap-southeast-2": {RULES_ECR_REPO_NAME: "184798709955"},
-    "ca-central-1": {RULES_ECR_REPO_NAME: "519511493484"},
-    "cn-north-1": {RULES_ECR_REPO_NAME: "618459771430"},
-    "cn-northwest-1": {RULES_ECR_REPO_NAME: "658757709296"},
-}
+from sagemaker import image_uris
+
+framework_name = "debugger"
 
 
 def get_rule_container_image_uri(region):
@@ -61,9 +38,7 @@ def get_rule_container_image_uri(region):
     Returns:
         str: Formatted image uri for the given region and the rule container type
     """
-    registry_id = SAGEMAKER_RULE_CONTAINERS_ACCOUNTS_MAP.get(region).get(RULES_ECR_REPO_NAME)
-    image_uri_prefix = get_ecr_image_uri_prefix(registry_id, region)
-    return "{}/{}:latest".format(image_uri_prefix, RULES_ECR_REPO_NAME)
+    return image_uris.retrieve(framework_name, region)
 
 
 class Rule(object):

src/sagemaker/deserializers.py

@@ -266,7 +266,8 @@ def deserialize(self, stream, content_type):
             list: A list of JSON serializable objects.
         """
         try:
-            lines = stream.read().rstrip().split("\n")
+            body = stream.read().decode("utf-8")
+            lines = body.rstrip().split("\n")
             return [json.loads(line) for line in lines]
         finally:
             stream.close()

src/sagemaker/estimator.py

@@ -29,11 +29,12 @@
 from sagemaker.debugger import DebuggerHookConfig
 from sagemaker.debugger import TensorBoardOutputConfig  # noqa: F401 # pylint: disable=unused-import
 from sagemaker.debugger import get_rule_container_image_uri
-from sagemaker.s3 import S3Uploader
+from sagemaker.deserializers import BytesDeserializer
+from sagemaker.s3 import S3Uploader, parse_s3_url
+from sagemaker.serializers import IdentitySerializer
 
 from sagemaker.fw_utils import (
     tar_and_upload_dir,
-    parse_s3_url,
     UploadedCode,
     validate_source_dir,
     _region_supports_debugger,
@@ -1341,16 +1342,14 @@ def create_model(
         role=None,
         image_uri=None,
         predictor_cls=None,
-        serializer=None,
-        deserializer=None,
-        content_type=None,
-        accept=None,
+        serializer=IdentitySerializer(),
+        deserializer=BytesDeserializer(),
         vpc_config_override=vpc_utils.VPC_CONFIG_DEFAULT,
         **kwargs
     ):
         """Create a model to deploy.
 
-        The serializer, deserializer, content_type, and accept arguments are only used to define a
+        The serializer and deserializer arguments are only used to define a
         default Predictor. They are ignored if an explicit predictor class is passed in.
         Other arguments are passed through to the Model class.
 
@@ -1362,17 +1361,12 @@ 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.
+            serializer (:class:`~sagemaker.serializers.BaseSerializer`): A
+                serializer object, used to encode data for an inference endpoint
+                (default: :class:`~sagemaker.serializers.IdentitySerializer`).
+            deserializer (:class:`~sagemaker.deserializers.BaseDeserializer`): A
+                deserializer object, used to decode data from an inference
+                endpoint (default: :class:`~sagemaker.deserializers.BytesDeserializer`).
             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 +1385,7 @@ 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, serializer, deserializer)
 
             predictor_cls = predict_wrapper
 
@@ -1418,7 +1412,7 @@ class Framework(EstimatorBase):
     such as training/deployment images and predictor instances.
     """
 
-    __framework_name__ = None
+    _framework_name = None
 
     LAUNCH_PS_ENV_NAME = "sagemaker_parameter_server_enabled"
     LAUNCH_MPI_ENV_NAME = "sagemaker_mpi_enabled"
@@ -1816,7 +1810,7 @@ def train_image(self):
         if self.image_uri:
             return self.image_uri
         return image_uris.retrieve(
-            self.__framework_name__,
+            self._framework_name,
             self.sagemaker_session.boto_region_name,
             instance_type=self.instance_type,
             version=self.framework_version,  # pylint: disable=no-member