Merge branch 'master' into master

Author: laurenyu

.github/PULL_REQUEST_TEMPLATE.md

@@ -8,11 +8,17 @@
 
 _Put an `x` in the boxes that apply. You can also fill these out after creating the PR. If you're unsure about any of them, don't hesitate to ask. We're here to help! This is simply a reminder of what we are going to look for before merging your pull request._
 
+#### General
+
 - [ ] I have read the [CONTRIBUTING](https://github.com/aws/sagemaker-python-sdk/blob/master/CONTRIBUTING.md) doc
 - [ ] I used the commit message format described in [CONTRIBUTING](https://github.com/aws/sagemaker-python-sdk/blob/master/CONTRIBUTING.md#committing-your-change)
 - [ ] I have used the regional endpoint when creating S3 and/or STS clients (if appropriate)
+- [ ] I have updated any necessary documentation, including [READMEs](https://github.com/aws/sagemaker-python-sdk/blob/master/README.rst) and [API docs](https://github.com/aws/sagemaker-python-sdk/tree/master/doc) (if appropriate)
+
+#### Tests
+
 - [ ] I have added tests that prove my fix is effective or that my feature works (if appropriate)
 - [ ] I have checked that my tests are not configured for a specific region or account (if appropriate)
-- [ ] I have updated any necessary documentation, including [READMEs](https://github.com/aws/sagemaker-python-sdk/blob/master/README.rst) and [API docs](https://github.com/aws/sagemaker-python-sdk/tree/master/doc) (if appropriate)
+- [ ] I have used [`unique_name_from_base`](https://github.com/aws/sagemaker-python-sdk/blob/master/src/sagemaker/utils.py#L77) to create resource names in integ tests (if appropriate)
 
 By submitting this pull request, I confirm that my contribution is made under the terms of the Apache 2.0 license.

README.rst

@@ -65,6 +65,7 @@ Table of Contents
 21. `SageMaker Workflow <#sagemaker-workflow>`__
 22. `SageMaker Autopilot <#sagemaker-autopilot>`__
 23. `Model Monitoring <#amazon-sagemaker-model-monitoring>`__
+24. `SageMaker Debugger <#amazon-sagemaker-debugger>`__
 
 
 Installing the SageMaker Python SDK
@@ -366,3 +367,13 @@ You can use Amazon SageMaker Model Monitoring to automatically detect concept dr
 For more information, see `Amazon SageMaker Model Monitoring`_.
 
 .. _Amazon SageMaker Model Monitoring: https://sagemaker.readthedocs.io/en/stable/amazon_sagemaker_model_monitoring.html
+
+Amazon SageMaker Debugger
+-------------------------
+
+You can use Amazon SageMaker Debugger to automatically detect anomalies while training your machine learning models.
+
+For more information, see `Amazon SageMaker Debugger`_.
+
+.. _Amazon SageMaker Debugger: https://sagemaker.readthedocs.io/en/stable/amazon_sagemaker_debugger.html
+

doc/amazon_sagemaker_debugger.rst

@@ -82,12 +82,11 @@ IAM role-based operator deployment
 Before you can deploy your operator using an IAM role, associate an OpenID Connect (OIDC) provider with your role to
 authenticate with the IAM service.
 
-Associate an OpenID Connect Provider to Your Instance
-^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
+Create an OpenID Connect Provider for Your Cluster
+^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
 
-Create an OIDC identity provider for your cluster. If your
-cluster is managed by EKS, then your cluster will already have an OIDC
-attached to it. 
+The following instruction will create and associate an OIDC provider
+with your EKS cluster.
 
 Set the local ``CLUSTER_NAME`` and \ ``AWS_REGION`` environment
 variables as follows:

doc/amazon_sagemaker_operators_for_kubernetes.rst

@@ -29,8 +29,13 @@ Download the sample YAML file for training using the following command:
 
     wget https://github.com/aws/amazon-sagemaker-operator-for-k8s/blob/master/samples/xgboost-mnist-trainingjob.yaml
 
-Edit the \ ``xgboost-mnist-trainingjob.yaml`` file to replace the ``roleArn`` parameter with your \ ``<sagemaker-execution-role>``, and \ ``outputPath`` with your S3 bucket that the Amazon SageMaker
-execution role has write access to. Apply the YAML file using the
+Edit the ``xgboost-mnist-trainingjob.yaml`` file to replace the ``roleArn`` parameter with your ``<sagemaker-execution-role>``, and ``outputPath`` with your S3 bucket that the Amazon SageMaker
+execution role has write access to. The ``roleArn`` must have permissions so that Amazon SageMaker
+can access Amazon S3, Amazon CloudWatch, and other services on your 
+behalf. For more information on creating an Amazon SageMaker
+ExecutionRole, see `Amazon SageMaker
+Roles <https://docs.aws.amazon.com/sagemaker/latest/dg/sagemaker-roles.html#sagemaker-roles-createtrainingjob-perms>`__.
+Apply the YAML file using the
 following command:
 
 ::

doc/amazon_sagemaker_operators_for_kubernetes_jobs.rst

@@ -208,3 +208,13 @@ You can use Amazon SageMaker Model Monitoring to automatically detect concept dr
     :maxdepth: 2
 
     amazon_sagemaker_model_monitoring
+
+*************************
+Amazon SageMaker Debugger
+*************************
+You can use Amazon SageMaker Debugger to automatically detect anomalies while training your machine learning models.
+
+.. toctree::
+    :maxdepth: 2
+
+    amazon_sagemaker_debugger

doc/index.rst

@@ -1048,6 +1048,15 @@ For more information, see `SageMaker Model Monitoring`_.
 
 .. _SageMaker Model Monitoring: https://github.com/aws/sagemaker-python-sdk/blob/master/doc/amazon_sagemaker_model_monitoring.rst
 
+******************
+SageMaker Debugger
+******************
+You can use Amazon SageMaker Debugger to automatically detect anomalies while training your machine learning models.
+
+For more information, see `SageMaker Debugger`_.
+
+.. _SageMaker Debugger: https://github.com/aws/sagemaker-python-sdk/blob/master/doc/amazon_sagemaker_debugger.rst
+
 ***
 FAQ
 ***

doc/overview.rst

@@ -116,7 +116,10 @@ When running your training script on SageMaker, it will have access to some pre-
 For more information on the runtime environment, including specific package versions, see `SageMaker Chainer Docker containers <#sagemaker-chainer-docker-containers>`__.
 
 If there are other packages you want to use with your script, you can include a ``requirements.txt`` file in the same directory as your training script to install other dependencies at runtime.
-A ``requirements.txt`` file is a text file that contains a list of items that are installed by using ``pip install``. You can also specify the version of an item to install.
+Both ``requirements.txt`` and your training script should be put in the same folder.
+You must specify this folder in ``source_dir`` argument when creating a Chainer estimator.
+A ``requirements.txt`` file is a text file that contains a list of items that are installed by using ``pip install``.
+You can also specify the version of an item to install.
 For information about the format of a ``requirements.txt`` file, see `Requirements Files <https://pip.pypa.io/en/stable/user_guide/#requirements-files>`__ in the pip documentation.
 
 Create an Estimator
@@ -637,4 +640,4 @@ SageMaker Chainer Docker containers
 
 You can visit the SageMaker Chainer containers repository here: https://github.com/aws/sagemaker-chainer-container
 
-For information about SageMaker Chainer Docker containers and their dependencies, see `SageMaker Chainer Docker containers <https://github.com/aws/sagemaker-python-sdk/tree/master/src/sagemaker/chainer#sagemaker-chainer-docker-containers>`_.
+For information about SageMaker Chainer Docker containers and their dependencies, see `SageMaker Chainer Docker containers <https://github.com/aws/sagemaker-python-sdk/tree/master/src/sagemaker/chainer#sagemaker-chainer-docker-containers>`_.

doc/using_chainer.rst

@@ -310,7 +310,10 @@ When running your training script on SageMaker, it will have access to some pre-
 For more information on the runtime environment, including specific package versions, see `SageMaker MXNet Containers <#sagemaker-mxnet-containers>`__.
 
 If there are other packages you want to use with your script, you can include a ``requirements.txt`` file in the same directory as your training script to install other dependencies at runtime.
-A ``requirements.txt`` file is a text file that contains a list of items that are installed by using ``pip install``. You can also specify the version of an item to install.
+Both ``requirements.txt`` and your training script should be put in the same folder.
+You must specify this folder in ``source_dir`` argument when creating an MXNet estimator.
+A ``requirements.txt`` file is a text file that contains a list of items that are installed by using ``pip install``.
+You can also specify the version of an item to install.
 For information about the format of a ``requirements.txt`` file, see `Requirements Files <https://pip.pypa.io/en/stable/user_guide/#requirements-files>`__ in the pip documentation.
 
 Create an Estimator

doc/using_mxnet.rst

@@ -131,7 +131,6 @@ The function of installing packages using ``requirements.txt`` is supported for
 
 A ``requirements.txt`` file is a text file that contains a list of items that are installed by using ``pip install``. You can also specify the version of an item to install. For information about the format of a ``requirements.txt`` file, see `Requirements Files <https://pip.pypa.io/en/stable/user_guide/#requirements-files>`__ in the pip documentation.
 
-
 Create an Estimator
 ===================
 

doc/using_pytorch.rst

@@ -114,11 +114,12 @@ When running your training script on SageMaker, it has access to some pre-instal
 For more information on the runtime environment, including specific package versions, see `SageMaker Scikit-learn Docker Container <https://github.com/aws/sagemaker-scikit-learn-container>`__.
 
 If there are other packages you want to use with your script, you can include a ``requirements.txt`` file in the same directory as your training script to install other dependencies at runtime.
-A ``requirements.txt`` file is a text file that contains a list of items that are installed by using ``pip install``. You can also specify the version of an item to install.
+Both ``requirements.txt`` and your training script should be put in the same folder.
+You must specify this folder in ``source_dir`` argument when creating a Scikit-learn estimator.
+A ``requirements.txt`` file is a text file that contains a list of items that are installed by using ``pip install``.
+You can also specify the version of an item to install.
 For information about the format of a ``requirements.txt`` file, see `Requirements Files <https://pip.pypa.io/en/stable/user_guide/#requirements-files>`__ in the pip documentation.
 
-
-
 Create an Estimator
 ===================
 

doc/using_sklearn.rst

@@ -201,6 +201,7 @@ def deploy(
         vpc_config=None,
         enable_network_isolation=False,
         model_kms_key=None,
+        predictor_cls=None,
     ):
         """Deploy a candidate to a SageMaker Inference Pipeline and return a Predictor
 
@@ -237,10 +238,15 @@ def deploy(
                 training cluster for distributed training. Default: False
             model_kms_key (str): KMS key ARN used to encrypt the repacked
                 model archive file if the model is repacked
+            predictor_cls (callable[string, sagemaker.session.Session]): A
+                function to call to create a predictor (default: None). If
+                specified, ``deploy()``  returns the result of invoking this
+                function on the created endpoint name.
 
         Returns:
-            callable[string, sagemaker.session.Session]: Invocation of
-            ``self.predictor_cls`` on the created endpoint name.
+            callable[string, sagemaker.session.Session] or ``None``:
+                If ``predictor_cls`` is specified, the invocation of ``self.predictor_cls`` on
+                the created endpoint name. Otherwise, ``None``.
         """
         if candidate is None:
             candidate_dict = self.best_candidate()
@@ -264,6 +270,7 @@ def deploy(
             vpc_config=vpc_config,
             enable_network_isolation=enable_network_isolation,
             model_kms_key=model_kms_key,
+            predictor_cls=predictor_cls,
         )
 
     def _check_problem_type_and_job_objective(self, problem_type, job_objective):
@@ -299,6 +306,7 @@ def _deploy_inference_pipeline(
         vpc_config=None,
         enable_network_isolation=False,
         model_kms_key=None,
+        predictor_cls=None,
     ):
         """Deploy a SageMaker Inference Pipeline.
 
@@ -329,6 +337,10 @@ def _deploy_inference_pipeline(
                 contains "SecurityGroupIds", "Subnets"
             model_kms_key (str): KMS key ARN used to encrypt the repacked
                 model archive file if the model is repacked
+            predictor_cls (callable[string, sagemaker.session.Session]): A
+                function to call to create a predictor (default: None). If
+                specified, ``deploy()``  returns the result of invoking this
+                function on the created endpoint name.
         """
         # construct Model objects
         models = []
@@ -352,6 +364,7 @@ def _deploy_inference_pipeline(
         pipeline = PipelineModel(
             models=models,
             role=self.role,
+            predictor_cls=predictor_cls,
             name=name,
             vpc_config=vpc_config,
             sagemaker_session=sagemaker_session or self.sagemaker_session,

src/sagemaker/automl/automl.py

@@ -17,3 +17,6 @@
 """Default Chainer version for when the framework version is not specified.
 This is no longer updated so as to not break existing workflows.
 """
+
+LATEST_VERSION = "5.0.0"
+"""The latest version of Chainer included in the SageMaker pre-built Docker images."""

src/sagemaker/chainer/defaults.py

@@ -22,7 +22,7 @@
     empty_framework_version_warning,
     python_deprecation_warning,
 )
-from sagemaker.chainer.defaults import CHAINER_VERSION
+from sagemaker.chainer.defaults import CHAINER_VERSION, LATEST_VERSION
 from sagemaker.chainer.model import ChainerModel
 from sagemaker.vpc_utils import VPC_CONFIG_DEFAULT
 
@@ -40,8 +40,7 @@ class Chainer(Framework):
     _process_slots_per_host = "sagemaker_process_slots_per_host"
     _additional_mpi_options = "sagemaker_additional_mpi_options"
 
-    LATEST_VERSION = "5.0.0"
-    """The latest version of Chainer included in the SageMaker pre-built Docker images."""
+    LATEST_VERSION = LATEST_VERSION
 
     def __init__(
         self,

src/sagemaker/chainer/estimator.py

@@ -16,9 +16,14 @@
 import logging
 
 import sagemaker
-from sagemaker.fw_utils import create_image_uri, model_code_key_prefix, python_deprecation_warning
+from sagemaker.fw_utils import (
+    create_image_uri,
+    model_code_key_prefix,
+    python_deprecation_warning,
+    empty_framework_version_warning,
+)
 from sagemaker.model import FrameworkModel, MODEL_SERVER_WORKERS_PARAM_NAME
-from sagemaker.chainer.defaults import CHAINER_VERSION
+from sagemaker.chainer.defaults import CHAINER_VERSION, LATEST_VERSION
 from sagemaker.predictor import RealTimePredictor, npy_serializer, numpy_deserializer
 
 logger = logging.getLogger("sagemaker")
@@ -61,7 +66,7 @@ def __init__(
         entry_point,
         image=None,
         py_version="py3",
-        framework_version=CHAINER_VERSION,
+        framework_version=None,
         predictor_cls=ChainerPredictor,
         model_server_workers=None,
         **kwargs
@@ -107,9 +112,11 @@ def __init__(
         )
         if py_version == "py2":
             logger.warning(python_deprecation_warning(self.__framework_name__))
+        if framework_version is None:
+            logger.warning(empty_framework_version_warning(CHAINER_VERSION, LATEST_VERSION))
 
         self.py_version = py_version
-        self.framework_version = framework_version
+        self.framework_version = framework_version or CHAINER_VERSION
         self.model_server_workers = model_server_workers
 
     def prepare_container_def(self, instance_type, accelerator_type=None):

src/sagemaker/chainer/model.py

@@ -473,7 +473,9 @@ def deploy(
                 kms_key=kms_key,
                 data_capture_config_dict=data_capture_config_dict,
             )
-            self.sagemaker_session.update_endpoint(self.endpoint_name, endpoint_config_name)
+            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,

src/sagemaker/model.py

@@ -243,7 +243,9 @@ def deploy(
                 kms_key=kms_key,
                 data_capture_config_dict=data_capture_config_dict,
             )
-            self.sagemaker_session.update_endpoint(self.endpoint_name, endpoint_config_name)
+            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,

src/sagemaker/multidatamodel.py

@@ -17,3 +17,6 @@
 """Default MXNet version for when the framework version is not specified.
 This is no longer updated so as to not break existing workflows.
 """
+
+LATEST_VERSION = "1.6.0"
+"""The latest version of MXNet included in the SageMaker pre-built Docker images."""

src/sagemaker/mxnet/defaults.py

@@ -23,7 +23,7 @@
     python_deprecation_warning,
     is_version_equal_or_higher,
 )
-from sagemaker.mxnet.defaults import MXNET_VERSION
+from sagemaker.mxnet.defaults import MXNET_VERSION, LATEST_VERSION
 from sagemaker.mxnet.model import MXNetModel
 from sagemaker.vpc_utils import VPC_CONFIG_DEFAULT
 
@@ -36,8 +36,7 @@ class MXNet(Framework):
     __framework_name__ = "mxnet"
     _LOWEST_SCRIPT_MODE_VERSION = ["1", "3"]
 
-    LATEST_VERSION = "1.6.0"
-    """The latest version of MXNet included in the SageMaker pre-built Docker images."""
+    LATEST_VERSION = LATEST_VERSION
 
     def __init__(
         self,

src/sagemaker/mxnet/estimator.py

@@ -18,9 +18,14 @@
 from pkg_resources import parse_version
 
 import sagemaker
-from sagemaker.fw_utils import create_image_uri, model_code_key_prefix, python_deprecation_warning
+from sagemaker.fw_utils import (
+    create_image_uri,
+    model_code_key_prefix,
+    python_deprecation_warning,
+    empty_framework_version_warning,
+)
 from sagemaker.model import FrameworkModel, MODEL_SERVER_WORKERS_PARAM_NAME
-from sagemaker.mxnet.defaults import MXNET_VERSION
+from sagemaker.mxnet.defaults import MXNET_VERSION, LATEST_VERSION
 from sagemaker.predictor import RealTimePredictor, json_serializer, json_deserializer
 
 logger = logging.getLogger("sagemaker")
@@ -62,7 +67,7 @@ def __init__(
         entry_point,
         image=None,
         py_version="py2",
-        framework_version=MXNET_VERSION,
+        framework_version=None,
         predictor_cls=MXNetPredictor,
         model_server_workers=None,
         **kwargs
@@ -109,9 +114,11 @@ def __init__(
 
         if py_version == "py2":
             logger.warning(python_deprecation_warning(self.__framework_name__))
+        if framework_version is None:
+            logger.warning(empty_framework_version_warning(MXNET_VERSION, LATEST_VERSION))
 
         self.py_version = py_version
-        self.framework_version = framework_version
+        self.framework_version = framework_version or MXNET_VERSION
         self.model_server_workers = model_server_workers
 
     def prepare_container_def(self, instance_type, accelerator_type=None):

src/sagemaker/mxnet/model.py

@@ -159,7 +159,9 @@ def deploy(
                 tags=tags,
                 data_capture_config_dict=data_capture_config_dict,
             )
-            self.sagemaker_session.update_endpoint(self.endpoint_name, endpoint_config_name)
+            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,

src/sagemaker/pipeline.py

@@ -19,4 +19,7 @@
 break existing workflows.
 """
 
+LATEST_VERSION = "1.3.1"
+"""The latest version of PyTorch included in the SageMaker pre-built Docker images."""
+
 PYTHON_VERSION = "py3"