Merge branch 'master' into processor-docs

Author: laurenyu

src/sagemaker/automl/automl.py

@@ -13,6 +13,7 @@
 """A class for SageMaker AutoML Jobs."""
 from __future__ import absolute_import
 
+import logging
 from six import string_types
 
 from sagemaker import Model, PipelineModel
@@ -21,6 +22,8 @@
 from sagemaker.session import Session
 from sagemaker.utils import name_from_base
 
+logger = logging.getLogger("sagemaker")
+
 
 class AutoML(object):
     """A class for creating and interacting with SageMaker AutoML jobs
@@ -78,16 +81,15 @@ def fit(self, inputs=None, wait=True, logs=True, job_name=None):
                 is stored. Or an AutoMLInput object. If a local path is provided, the dataset will
                 be uploaded to an S3 location.
             wait (bool): Whether the call should wait until the job completes (default: True).
-            logs (bool): Whether to show the logs produced by the job.
-                Only meaningful when wait is True (default: True).
+            logs (bool): Whether to show the logs produced by the job. Only meaningful when wait
+                is True (default: True). if ``wait`` is False, ``logs`` will be set to False as
+                well.
             job_name (str): Training job name. If not specified, the estimator generates
                 a default job name, based on the training image name and current timestamp.
         """
-        if logs and not wait:
-            raise ValueError(
-                """Logs can only be shown if wait is set to True.
-                Please either set wait to True or set logs to False."""
-            )
+        if not wait and logs:
+            logs = False
+            logger.warning("Setting logs to False. logs is only meaningful when wait is True.")
 
         # upload data for users if provided local path
         # validations are done in _Job._format_inputs_to_input_config

src/sagemaker/fw_utils.py

@@ -49,8 +49,8 @@
     "Please set the argument \"py_version='py3'\" to use the Python 3 {framework} image."
 )
 PARAMETER_SERVER_MULTI_GPU_WARNING = (
-    "You have selected a multi-GPU training instance type. "
-    "You have also enabled parameter server for distributed training. "
+    "If you have selected a multi-GPU training instance type, "
+    "and have also enabled parameter server for distributed training. "
     "Distributed training with the default parameter server configuration will not "
     "fully leverage all GPU cores; the parameter server will be configured to run "
     "only one worker per host regardless of the number of GPUs."
@@ -625,9 +625,9 @@ def warn_if_parameter_server_with_multi_gpu(training_instance_type, distribution
         return
 
     is_multi_gpu_instance = (
-        training_instance_type.split(".")[1].startswith("p")
-        and training_instance_type not in SINGLE_GPU_INSTANCE_TYPES
-    )
+        training_instance_type == "local_gpu"
+        or training_instance_type.split(".")[1].startswith("p")
+    ) and training_instance_type not in SINGLE_GPU_INSTANCE_TYPES
 
     ps_enabled = "parameter_server" in distributions and distributions["parameter_server"].get(
         "enabled", False

src/sagemaker/workflow/airflow.py

@@ -181,6 +181,9 @@ def training_base_config(estimator, inputs=None, job_name=None, mini_batch_size=
     if job_config["vpc_config"] is not None:
         train_config["VpcConfig"] = job_config["vpc_config"]
 
+    if estimator.train_use_spot_instances:
+        train_config["EnableManagedSpotTraining"] = True
+
     if estimator.hyperparameters() is not None:
         hyperparameters = {str(k): str(v) for (k, v) in estimator.hyperparameters().items()}
 

tests/unit/sagemaker/automl/test_auto_ml.py

@@ -294,6 +294,17 @@ def test_auto_ml_only_one_of_problem_type_and_job_objective_provided(sagemaker_s
         )
 
 
+@patch("sagemaker.automl.automl.AutoMLJob.start_new")
+def test_auto_ml_fit_set_logs_to_false(start_new, sagemaker_session, caplog):
+    auto_ml = AutoML(
+        role=ROLE, target_attribute_name=TARGET_ATTRIBUTE_NAME, sagemaker_session=sagemaker_session
+    )
+    inputs = DEFAULT_S3_INPUT_DATA
+    auto_ml.fit(inputs, job_name=JOB_NAME, wait=False, logs=True)
+    start_new.wait.assert_not_called()
+    assert "Setting logs to False. logs is only meaningful when wait is True." in caplog.text
+
+
 def test_auto_ml_additional_optional_params(sagemaker_session):
     auto_ml = AutoML(
         role=ROLE,

tests/unit/test_airflow.py

@@ -105,6 +105,7 @@ def test_byo_training_config_all_args(sagemaker_session):
         model_uri="{{ model_uri }}",
         model_channel_name="{{ model_chanel }}",
         sagemaker_session=sagemaker_session,
+        train_use_spot_instances=True,
     )
 
     byo.set_hyperparameters(epochs=32, feature_dim=1024, mini_batch_size=256)
@@ -155,6 +156,7 @@ def test_byo_training_config_all_args(sagemaker_session):
             "Subnets": ["{{ subnet }}"],
             "SecurityGroupIds": ["{{ security_group_ids }}"],
         },
+        "EnableManagedSpotTraining": True,
         "HyperParameters": {"epochs": "32", "feature_dim": "1024", "mini_batch_size": "256"},
         "Tags": [{"{{ key }}": "{{ value }}"}],
     }

tests/unit/test_fw_utils.py

@@ -1272,3 +1272,13 @@ def test_warn_if_parameter_server_with_multi_gpu(caplog):
         training_instance_type=train_instance_type, distributions=distributions
     )
     assert fw_utils.PARAMETER_SERVER_MULTI_GPU_WARNING in caplog.text
+
+
+def test_warn_if_parameter_server_with_local_multi_gpu(caplog):
+    train_instance_type = "local_gpu"
+    distributions = {"parameter_server": {"enabled": True}}
+
+    fw_utils.warn_if_parameter_server_with_multi_gpu(
+        training_instance_type=train_instance_type, distributions=distributions
+    )
+    assert fw_utils.PARAMETER_SERVER_MULTI_GPU_WARNING in caplog.text