Merge branch 'master' into f/sm-local-selinux

Author: sodre

CHANGELOG.md

@@ -1,5 +1,61 @@
 # Changelog
 
+## v2.154.0 (2023-05-11)
+
+### Features
+
+ * Add integ tests for remote_function, auto_capture functionality
+ * jumpstart model estimator classes
+
+### Bug Fixes and Other Changes
+
+ * integs - pytorch transformer deps and add test retry
+ * adding .lower() so new Pandas dtypes will match the type lookup.
+ * Pass KMS value to create processing job
+
+## v2.153.0 (2023-05-09)
+
+### Features
+
+ * Support npz archives in NumpyDeserializer
+ * Add FasterTransformer DJL support
+ * support for Sample Weights for SageMaker Autopilot
+
+### Bug Fixes and Other Changes
+
+ * retry is_run assertion
+ * Avoid 'AttributeError' for endpoint_name, if deploy() is not yet called
+ * Fix LambdaStep Creation
+ * Fix error when instance_count>1 in remote_function
+ * Remove deprecated update_endpoint from deploy() args in TensorFlowModel
+ * Update DJL deepspeed and fastertransformer DLC image uris
+ * remote_function python version mismatch issue
+
+## v2.152.0 (2023-05-04)
+
+### Features
+
+ * add support for lineage visualization using pyvis
+ * Expose Experiment class publicly
+ * PyTorch 1.13 release
+
+### Bug Fixes and Other Changes
+
+ * Change data_type argument to dtype to keep consistent with D…
+ * Skip edge test
+ * make RemoteExecutor context manager non-blocking on pending futures
+ * Add inferentia2 DLC images for djl framework
+ * Fix typo in using_pytorch.rst
+ * Unable to attach estimator to training job when KeepAlivePeriodInSeconds specified
+ * update LMI container image
+ * Update Clarify SHAPConfig baseline to allow JSON structures
+
+### Documentation Changes
+
+ * Fix broken link in DJL SageMaker docs
+ * currency update for the SageMaker data parallelism lib
+ * SM model parallel library v1.15.0 release note
+
 ## v2.151.0 (2023-04-27)
 
 ### Features

VERSION

@@ -1 +1 @@
-2.151.1.dev0
+2.154.1.dev0

doc/frameworks/djl/sagemaker.djl_inference.rst

@@ -26,6 +26,14 @@ HuggingFaceAccelerateModel
     :undoc-members:
     :show-inheritance:
 
+FasterTransformerModel
+---------------------------
+
+.. autoclass:: sagemaker.djl_inference.model.FasterTransformerModel
+    :members:
+    :undoc-members:
+    :show-inheritance:
+
 DJLPredictor
 ---------------------------
 

doc/frameworks/djl/using_djl.rst

@@ -23,7 +23,7 @@ With the SageMaker Python SDK, you can use DJL Serving to host models that have
 These can either be models you have trained/fine-tuned yourself, or models available publicly from the HuggingFace Hub.
 DJL Serving in the SageMaker Python SDK supports hosting models for the popular HuggingFace NLP tasks, as well as Stable Diffusion.
 
-You can either deploy your model using DeepSpeed or HuggingFace Accelerate, or let DJL Serving determine the best backend based on your model architecture and configuration.
+You can either deploy your model using DeepSpeed, FasterTransformer, or HuggingFace Accelerate, or let DJL Serving determine the best backend based on your model architecture and configuration.
 
 .. code:: python
 
@@ -63,11 +63,23 @@ If you want to use a specific backend, then you can create an instance of the co
         number_of_partitions=2, # number of gpus to partition the model across
     )
 
+    # Create a model using the FasterTransformer backend
+
+    fastertransformer_model = FasterTransformerModel(
+        "s3://my_bucket/my_saved_model_artifacts/", # This can also be a HuggingFace Hub model id
+        "my_sagemaker_role",
+        data_type="fp16",
+        task="text-generation",
+        tensor_parallel_degree=2, # number of gpus to partition the model across
+    )
+
     # Deploy the model to an Amazon SageMaker Endpoint and get a Predictor
     deepspeed_predictor = deepspeed_model.deploy("ml.g5.12xlarge",
                                                  initial_instance_count=1)
     hf_accelerate_predictor = hf_accelerate_model.deploy("ml.g5.12xlarge",
                                                          initial_instance_count=1)
+    fastertransformer_predictor = fastertransformer_model.deploy("ml.g5.12xlarge",
+                                                                 initial_instance_count=1)
 
 Regardless of which way you choose to create your model, a ``Predictor`` object is returned. You can use this ``Predictor``
 to do inference on the endpoint hosting your DJLModel.

setup.py

@@ -48,7 +48,7 @@ def read_requirements(filename):
 # Declare minimal set for installation
 required_packages = [
     "attrs>=20.3.0,<23",
-    "boto3>=1.26.28,<2.0",
+    "boto3>=1.26.131,<2.0",
     "cloudpickle==2.2.1",
     "google-pasta",
     "numpy>=1.9.0,<2.0",

src/sagemaker/accept_types.py

@@ -0,0 +1,103 @@
+# Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved.
+#
+# Licensed under the Apache License, Version 2.0 (the "License"). You
+# may not use this file except in compliance with the License. A copy of
+# the License is located at
+#
+#     http://aws.amazon.com/apache2.0/
+#
+# or in the "license" file accompanying this file. This file is
+# distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF
+# ANY KIND, either express or implied. See the License for the specific
+# language governing permissions and limitations under the License.
+"""This module is for SageMaker accept types."""
+from __future__ import absolute_import
+from typing import List, Optional
+
+from sagemaker.jumpstart import artifacts, utils as jumpstart_utils
+
+
+def retrieve_options(
+    region: Optional[str] = None,
+    model_id: Optional[str] = None,
+    model_version: Optional[str] = None,
+    tolerate_vulnerable_model: bool = False,
+    tolerate_deprecated_model: bool = False,
+) -> List[str]:
+    """Retrieves the supported accept types for the model matching the given arguments.
+
+    Args:
+        region (str): The AWS Region for which to retrieve the supported accept types.
+            Defaults to ``None``.
+        model_id (str): The model ID of the model for which to
+            retrieve the supported accept types. (Default: None).
+        model_version (str): The version of the model for which to retrieve the
+            supported accept types. (Default: None).
+        tolerate_vulnerable_model (bool): True if vulnerable versions of model
+            specifications should be tolerated (exception not raised). If False, raises an
+            exception if the script used by this version of the model has dependencies with known
+            security vulnerabilities. (Default: False).
+        tolerate_deprecated_model (bool): True if deprecated models should be tolerated
+            (exception not raised). False if these models should raise an exception.
+            (Default: False).
+    Returns:
+        list: The supported accept types to use for the model.
+
+    Raises:
+        ValueError: If the combination of arguments specified is not supported.
+    """
+    if not jumpstart_utils.is_jumpstart_model_input(model_id, model_version):
+        raise ValueError(
+            "Must specify JumpStart `model_id` and `model_version` when retrieving accept types."
+        )
+
+    return artifacts._retrieve_supported_accept_types(
+        model_id,
+        model_version,
+        region,
+        tolerate_vulnerable_model,
+        tolerate_deprecated_model,
+    )
+
+
+def retrieve_default(
+    region: Optional[str] = None,
+    model_id: Optional[str] = None,
+    model_version: Optional[str] = None,
+    tolerate_vulnerable_model: bool = False,
+    tolerate_deprecated_model: bool = False,
+) -> str:
+    """Retrieves the default accept type for the model matching the given arguments.
+
+    Args:
+        region (str): The AWS Region for which to retrieve the default accept type.
+            Defaults to ``None``.
+        model_id (str): The model ID of the model for which to
+            retrieve the default accept type. (Default: None).
+        model_version (str): The version of the model for which to retrieve the
+            default accept type. (Default: None).
+        tolerate_vulnerable_model (bool): True if vulnerable versions of model
+            specifications should be tolerated (exception not raised). If False, raises an
+            exception if the script used by this version of the model has dependencies with known
+            security vulnerabilities. (Default: False).
+        tolerate_deprecated_model (bool): True if deprecated models should be tolerated
+            (exception not raised). False if these models should raise an exception.
+            (Default: False).
+    Returns:
+        str: The default accept type to use for the model.
+
+    Raises:
+        ValueError: If the combination of arguments specified is not supported.
+    """
+    if not jumpstart_utils.is_jumpstart_model_input(model_id, model_version):
+        raise ValueError(
+            "Must specify JumpStart `model_id` and `model_version` when retrieving accept types."
+        )
+
+    return artifacts._retrieve_default_accept_type(
+        model_id,
+        model_version,
+        region,
+        tolerate_vulnerable_model,
+        tolerate_deprecated_model,
+    )

src/sagemaker/automl/automl.py

@@ -49,6 +49,7 @@ def __init__(
         channel_type=None,
         content_type=None,
         s3_data_type=None,
+        sample_weight_attribute_name=None,
     ):
         """Convert an S3 Uri or a list of S3 Uri to an AutoMLInput object.
 
@@ -67,13 +68,16 @@ def __init__(
                 The content type of the data from the input source.
             s3_data_type (str, PipelineVariable): The data type for S3 data source.
                 Valid values: ManifestFile or S3Prefix.
+            sample_weight_attribute_name (str, PipelineVariable):
+                the name of the dataset column representing sample weights
         """
         self.inputs = inputs
         self.target_attribute_name = target_attribute_name
         self.compression = compression
         self.channel_type = channel_type
         self.content_type = content_type
         self.s3_data_type = s3_data_type
+        self.sample_weight_attribute_name = sample_weight_attribute_name
 
     def to_request_dict(self):
         """Generates a request dictionary using the parameters provided to the class."""
@@ -96,6 +100,8 @@ def to_request_dict(self):
                 input_entry["ContentType"] = self.content_type
             if self.s3_data_type is not None:
                 input_entry["DataSource"]["S3DataSource"]["S3DataType"] = self.s3_data_type
+            if self.sample_weight_attribute_name is not None:
+                input_entry["SampleWeightAttributeName"] = self.sample_weight_attribute_name
             auto_ml_input.append(input_entry)
         return auto_ml_input
 
@@ -129,6 +135,7 @@ def __init__(
         mode: Optional[str] = None,
         auto_generate_endpoint_name: Optional[bool] = None,
         endpoint_name: Optional[str] = None,
+        sample_weight_attribute_name: str = None,
     ):
         """Initialize the an AutoML object.
 
@@ -179,6 +186,8 @@ def __init__(
                 model deployment if the endpoint name is not generated automatically.
                 Specify the endpoint_name if and only if
                 auto_generate_endpoint_name is set to False
+            sample_weight_attribute_name (str): The name of dataset column representing
+                sample weights.
 
         Returns:
             AutoML object.
@@ -234,6 +243,7 @@ def __init__(
         )
 
         self._check_problem_type_and_job_objective(self.problem_type, self.job_objective)
+        self.sample_weight_attribute_name = sample_weight_attribute_name
 
     @runnable_by_pipeline
     def fit(self, inputs=None, wait=True, logs=True, job_name=None):
@@ -342,6 +352,9 @@ def attach(cls, auto_ml_job_name, sagemaker_session=None):
                 "AutoGenerateEndpointName", False
             ),
             endpoint_name=auto_ml_job_desc.get("ModelDeployConfig", {}).get("EndpointName"),
+            sample_weight_attribute_name=auto_ml_job_desc["InputDataConfig"][0].get(
+                "SampleWeightAttributeName", None
+            ),
         )
         amlj.current_job_name = auto_ml_job_name
         amlj.latest_auto_ml_job = auto_ml_job_name  # pylint: disable=W0201
@@ -867,6 +880,7 @@ def _load_config(cls, inputs, auto_ml, expand_role=True, validate_uri=True):
                 auto_ml.target_attribute_name,
                 auto_ml.content_type,
                 auto_ml.s3_data_type,
+                auto_ml.sample_weight_attribute_name,
             )
         output_config = _Job._prepare_output_config(auto_ml.output_path, auto_ml.output_kms_key)
 
@@ -932,6 +946,7 @@ def _format_inputs_to_input_config(
         target_attribute_name=None,
         content_type=None,
         s3_data_type=None,
+        sample_weight_attribute_name=None,
     ):
         """Convert inputs to AutoML InputDataConfig.
 
@@ -961,6 +976,8 @@ def _format_inputs_to_input_config(
                 channel["ContentType"] = content_type
             if s3_data_type is not None:
                 channel["DataSource"]["S3DataSource"]["S3DataType"] = s3_data_type
+            if sample_weight_attribute_name is not None:
+                channel["SampleWeightAttributeName"] = sample_weight_attribute_name
             channels.append(channel)
         elif isinstance(inputs, list):
             for input_entry in inputs:
@@ -974,6 +991,8 @@ def _format_inputs_to_input_config(
                     channel["ContentType"] = content_type
                 if s3_data_type is not None:
                     channel["DataSource"]["S3DataSource"]["S3DataType"] = s3_data_type
+                if sample_weight_attribute_name is not None:
+                    channel["SampleWeightAttributeName"] = sample_weight_attribute_name
                 channels.append(channel)
         else:
             msg = (