Merge branch 'master' into feat/jumpstart-model-id-list

Author: evakravi

.githooks/pre-push

@@ -12,5 +12,5 @@ start_time=`date +%s`
 tox -e sphinx,doc8 --parallel all
 ./ci-scripts/displaytime.sh 'sphinx,doc8' $start_time
 start_time=`date +%s`
-tox -e py36,py37,py38 --parallel all -- tests/unit
-./ci-scripts/displaytime.sh 'py36,py37,py38 unit' $start_time
+tox -e py36,py37,py38,py39 --parallel all -- tests/unit
+./ci-scripts/displaytime.sh 'py36,py37,py38,py39 unit' $start_time

CHANGELOG.md

@@ -1,5 +1,69 @@
 # Changelog
 
+## v2.84.0 (2022-04-07)
+
+### Features
+
+ * dependabot integ - move all deps to requirements.txt
+ * add xgboost framework version 1.5-1
+
+## v2.83.0 (2022-04-04)
+
+### Features
+
+ * Hugging Face Transformers 4.17 for TF 2.6
+
+### Bug Fixes and Other Changes
+
+ * IOC image version select issue
+
+## v2.82.2 (2022-04-01)
+
+### Bug Fixes and Other Changes
+
+ * Revert "fix: Fix Pipeline variables related customer issues (#2959)"
+ * Refactor repack_model script injection, fixes tar.gz error
+
+## v2.82.1 (2022-03-31)
+
+### Bug Fixes and Other Changes
+
+ * Update Inferentia Image URI Config
+ * Fix Pipeline variables related customer issues
+ * more logging info for static pipeline test data setup
+
+## v2.82.0 (2022-03-30)
+
+### Features
+
+ * pluggable instance fallback mechanism, add CapacityError
+ * support passing Env Vars to local mode training
+
+## v2.81.1 (2022-03-29)
+
+### Bug Fixes and Other Changes
+
+ * Update black-check version, add support for Spark 3.1 Processing
+
+## v2.81.0 (2022-03-26)
+
+### Features
+
+ * Retrieve data configuration
+ * enable EnableInterContainerTrafficEncryption for model monitoring
+ * Hugging Face Transformers 4.17 for PT 1.10
+
+### Bug Fixes and Other Changes
+
+ * remove `new` from serverless
+ * temporarily skip tests impacted by data inconsistency
+ * Implement override solution for pipeline variables
+
+### Documentation Changes
+
+ * add documentation for image_uri serverless use case
+ * minor fixes for smddp 1.4.0 doc
+
 ## v2.80.0 (2022-03-18)
 
 ### Features

MANIFEST.in

@@ -1,6 +1,7 @@
 recursive-include src/sagemaker *.py
 
 include src/sagemaker/image_uri_config/*.json
+recursive-include requirements *
 
 include VERSION
 include LICENSE.txt

VERSION

@@ -1 +1 @@
-2.80.1.dev0
+2.84.1.dev0

doc/_static/js/analytics.js

@@ -0,0 +1,4 @@
+$(document).ready( function () {
+    $('table.datatable').DataTable();
+    $('a.external').attr('target', '_blank');
+} );

doc/_static/js/datatable.js

@@ -67,6 +67,11 @@ def create_jumpstart_model_table():
     We highly suggest pinning an exact model version however.\n
     """
     )
+    file_content.append(
+        """
+    Each model id is linked to an external page that describes the model.\n
+    """
+    )
     file_content.append("\n")
     file_content.append(".. list-table:: Available Models\n")
     file_content.append("   :widths: 50 20 20 20\n")
@@ -80,7 +85,7 @@ def create_jumpstart_model_table():
 
     for model in sdk_manifest_top_versions_for_models.values():
         model_spec = get_jumpstart_sdk_spec(model["spec_key"])
-        file_content.append("   * - {}\n".format(model["model_id"]))
+        file_content.append("   * - `{} <{}>`_\n".format(model_spec["model_id"], model_spec["url"]))
         file_content.append("     - {}\n".format(model_spec["training_supported"]))
         file_content.append("     - {}\n".format(model["version"]))
         file_content.append("     - {}\n".format(model["min_version"]))

doc/doc_utils/jumpstart_doc_utils.py

@@ -1226,6 +1226,28 @@ to configure or manage the underlying infrastructure. After you trained a model,
 Serverless endpoint and then invoke the endpoint with the model to get inference results back. More information about
 SageMaker Serverless Inference can be found in the `AWS documentation <https://docs.aws.amazon.com/sagemaker/latest/dg/serverless-endpoints.html>`__.
 
+For using SageMaker Serverless Inference, you can either use SageMaker-provided container or Bring Your Own Container model.
+A step by step example for using Serverless Inference with MXNet image :
+
+Firstly, create MXNet model
+
+.. code:: python
+
+    from sagemaker.mxnet import MXNetModel
+    from sagemaker.serverless import ServerlessInferenceConfig
+    import sagemaker
+
+    role = sagemaker.get_execution_role()
+
+    # create MXNet Model Class
+    model = MXNetModel(
+        model_data="s3://my_bucket/pretrained_model/model.tar.gz", # path to your trained sagemaker model
+        role=role, # iam role with permissions to create an Endpoint
+        entry_point="inference.py",
+        py_version="py3", # Python version
+        framework_version="1.6.0", # MXNet framework version
+    )
+
 To deploy serverless endpoint, you will need to create a ``ServerlessInferenceConfig``.
 If you create ``ServerlessInferenceConfig`` without specifying its arguments, the default ``MemorySizeInMB`` will be **2048** and
 the default ``MaxConcurrency`` will be **5** :
@@ -1235,14 +1257,14 @@ the default ``MaxConcurrency`` will be **5** :
     from sagemaker.serverless import ServerlessInferenceConfig
 
     # Create an empty ServerlessInferenceConfig object to use default values
-    serverless_config = new ServerlessInferenceConfig()
+    serverless_config = ServerlessInferenceConfig()
 
 Or you can specify ``MemorySizeInMB`` and ``MaxConcurrency`` in ``ServerlessInferenceConfig`` (example shown below):
 
 .. code:: python
 
     # Specify MemorySizeInMB and MaxConcurrency in the serverless config object
-    serverless_config = new ServerlessInferenceConfig(
+    serverless_config = ServerlessInferenceConfig(
       memory_size_in_mb=4096,
       max_concurrency=10,
     )
@@ -1254,6 +1276,13 @@ Then use the ``ServerlessInferenceConfig`` in the estimator's ``deploy()`` metho
     # Deploys the model that was generated by fit() to a SageMaker serverless endpoint
     serverless_predictor = estimator.deploy(serverless_inference_config=serverless_config)
 
+Or directly using model's ``deploy()`` method to deploy a serverless endpoint:
+
+.. code:: python
+
+    # Deploys the model to a SageMaker serverless endpoint
+    serverless_predictor = model.deploy(serverless_inference_config=serverless_config)
+
 After deployment is complete, you can use predictor's ``predict()`` method to invoke the serverless endpoint just like
 real-time endpoints:
 

doc/overview.rst

@@ -0,0 +1,4 @@
+urllib3==1.26.8
+docker-compose==1.29.2
+docker~=5.0.0
+PyYAML==5.4.1

requirements/extras/local_requirements.txt

@@ -0,0 +1 @@
+scipy==1.5.4

requirements/extras/scipy_requirements.txt

@@ -0,0 +1,20 @@
+tox==3.24.5
+flake8==4.0.1
+pytest==6.0.2
+pytest-cov==3.0.0
+pytest-rerunfailures==10.2
+pytest-timeout==2.1.0
+pytest-xdist==2.4.0
+coverage>=5.2, <6.2
+mock==4.0.3
+contextlib2==21.6.0
+awslogs==0.14.0
+black==22.3.0
+stopit==1.1.2
+apache-airflow==2.2.4
+apache-airflow-providers-amazon==3.0.0
+attrs==20.3.0
+fabric==2.6.0
+requests==2.27.1
+sagemaker-experiments==0.1.35
+Jinja2==3.0.3

requirements/extras/test_requirements.txt

@@ -0,0 +1,2 @@
+doc8==0.10.1
+Pygments==2.11.2

requirements/tox/doc8_requirements.txt

@@ -0,0 +1 @@
+pydocstyle==6.1.1

requirements/tox/docstyle_requirements.txt

@@ -0,0 +1,2 @@
+flake8==4.0.1
+flake8-future-import==0.4.6

requirements/tox/flake8_requirements.txt

@@ -0,0 +1 @@
+mypy==0.942

requirements/tox/mypy_requirements.txt

@@ -0,0 +1 @@
+pydocstyle==6.1.1

requirements/tox/pydocstyle_requirements.txt

@@ -0,0 +1,2 @@
+pylint==2.6.2
+astroid==2.4.2

requirements/tox/pylint_requirements.txt

@@ -0,0 +1,2 @@
+pyenchant==3.2.2
+pylint==2.6.2

requirements/tox/spelling_requirements.txt

@@ -0,0 +1 @@
+twine==3.8.0

requirements/tox/twine_requirements.txt

@@ -16,7 +16,7 @@
 import os
 from glob import glob
 
-from setuptools import setup, find_packages
+from setuptools import find_packages, setup
 
 
 def read(fname):
@@ -31,6 +31,20 @@ def read_version():
     return read("VERSION").strip()
 
 
+def read_requirements(filename):
+    """Reads requirements file which lists package dependencies.
+
+    Args:
+        filename: type(str) Relative file path of requirements.txt file
+
+    Returns:
+        list of dependencies extracted from file
+    """
+    with open(os.path.abspath(filename)) as fp:
+        deps = [line.strip() for line in fp.readlines()]
+    return deps
+
+
 # Declare minimal set for installation
 required_packages = [
     "attrs==20.3.0",
@@ -47,42 +61,15 @@ def read_version():
 ]
 
 # Specific use case dependencies
+# Keep format of *_requirements.txt to be tracked by dependabot
 extras = {
-    "local": [
-        "urllib3==1.26.8",
-        "docker-compose==1.29.2",
-        "docker~=5.0.0",
-        "PyYAML==5.4.1",  # PyYAML version has to match docker-compose requirements
-    ],
-    "scipy": ["scipy==1.5.4"],
+    "local": read_requirements("requirements/extras/local_requirements.txt"),
+    "scipy": read_requirements("requirements/extras/scipy_requirements.txt"),
 }
 # Meta dependency groups
 extras["all"] = [item for group in extras.values() for item in group]
 # Tests specific dependencies (do not need to be included in 'all')
-extras["test"] = (
-    [
-        extras["all"],
-        "tox==3.24.5",
-        "flake8==4.0.1",
-        "pytest==6.0.2",
-        "pytest-cov==3.0.0",
-        "pytest-rerunfailures==10.2",
-        "pytest-timeout==2.1.0",
-        "pytest-xdist==2.4.0",
-        "coverage>=5.2, <6.2",
-        "mock==4.0.3",
-        "contextlib2==21.6.0",
-        "awslogs==0.14.0",
-        "black==22.1.0",
-        "stopit==1.1.2",
-        "apache-airflow==2.2.3",
-        "apache-airflow-providers-amazon==3.0.0",
-        "attrs==20.3.0",
-        "fabric==2.6.0",
-        "requests==2.27.1",
-        "sagemaker-experiments==0.1.35",
-    ],
-)
+extras["test"] = (extras["all"] + read_requirements("requirements/extras/test_requirements.txt"),)
 
 setup(
     name="sagemaker",

setup.py

@@ -99,10 +99,11 @@ def __init__(
                 file which should be executed as the entry point to model
                 hosting. If ``source_dir`` is specified, then ``entry_point``
                 must point to a file located at the root of ``source_dir``.
-            image_uri (str): A Docker image URI (default: None). If not specified, a
-                default image for Chainer will be used. If ``framework_version``
-                or ``py_version`` are ``None``, then ``image_uri`` is required. If
-                also ``None``, then a ``ValueError`` will be raised.
+            image_uri (str): A Docker image URI (default: None). If not specified,
+                a default image for Chainer will be used.
+                If ``framework_version`` or ``py_version``
+                are ``None``, then ``image_uri`` is required. If ``image_uri`` is also ``None``,
+                then a ``ValueError`` will be raised.
             framework_version (str): Chainer version you want to use for
                 executing your model training code. Defaults to ``None``. Required
                 unless ``image_uri`` is provided.
@@ -139,7 +140,9 @@ def __init__(
 
         self.model_server_workers = model_server_workers
 
-    def prepare_container_def(self, instance_type=None, accelerator_type=None):
+    def prepare_container_def(
+        self, instance_type=None, accelerator_type=None, serverless_inference_config=None
+    ):
         """Return a container definition with framework configuration set in model environment.
 
         Args:
@@ -148,21 +151,27 @@ def prepare_container_def(self, instance_type=None, accelerator_type=None):
             accelerator_type (str): The Elastic Inference accelerator type to
                 deploy to the instance for loading and making inferences to the
                 model. For example, 'ml.eia1.medium'.
+            serverless_inference_config (sagemaker.serverless.ServerlessInferenceConfig):
+                Specifies configuration related to serverless endpoint. Instance type is
+                not provided in serverless inference. So this is used to find image URIs.
 
         Returns:
             dict[str, str]: A container definition object usable with the
             CreateModel API.
         """
         deploy_image = self.image_uri
         if not deploy_image:
-            if instance_type is None:
+            if instance_type is None and serverless_inference_config is None:
                 raise ValueError(
                     "Must supply either an instance type (for choosing CPU vs GPU) or an image URI."
                 )
 
             region_name = self.sagemaker_session.boto_session.region_name
             deploy_image = self.serving_image_uri(
-                region_name, instance_type, accelerator_type=accelerator_type
+                region_name,
+                instance_type,
+                accelerator_type=accelerator_type,
+                serverless_inference_config=serverless_inference_config,
             )
 
         deploy_key_prefix = model_code_key_prefix(self.key_prefix, self.name, deploy_image)
@@ -174,13 +183,18 @@ def prepare_container_def(self, instance_type=None, accelerator_type=None):
             deploy_env[MODEL_SERVER_WORKERS_PARAM_NAME.upper()] = str(self.model_server_workers)
         return sagemaker.container_def(deploy_image, self.model_data, deploy_env)
 
-    def serving_image_uri(self, region_name, instance_type, accelerator_type=None):
+    def serving_image_uri(
+        self, region_name, instance_type, accelerator_type=None, serverless_inference_config=None
+    ):
         """Create a URI for the serving image.
 
         Args:
             region_name (str): AWS region where the image is uploaded.
             instance_type (str): SageMaker instance type. Used to determine device type
                 (cpu/gpu/family-specific optimized).
+            serverless_inference_config (sagemaker.serverless.ServerlessInferenceConfig):
+                Specifies configuration related to serverless endpoint. Instance type is
+                not provided in serverless inference. So this is used to determine device type.
 
         Returns:
             str: The appropriate image URI based on the given parameters.
@@ -194,4 +208,5 @@ def serving_image_uri(self, region_name, instance_type, accelerator_type=None):
             instance_type=instance_type,
             accelerator_type=accelerator_type,
             image_scope="inference",
+            serverless_inference_config=serverless_inference_config,
         )