Add SageMaker integ test for hyperparameter tuning model_dir logic (#183)

Author: laurenyu

docker/1.12.0/Dockerfile.gpu

@@ -107,12 +107,11 @@ WORKDIR /
 ARG framework_installable
 ARG framework_support_installable=sagemaker_tensorflow_container-2.0.0.tar.gz
 
-COPY $framework_installable tensorflow-1.12.0-py2.py3-none-any.whl 
+COPY $framework_installable tensorflow-1.12.0-py2.py3-none-any.whl
 COPY $framework_support_installable .
 
 RUN pip install --no-cache-dir -U \
     keras==2.2.4 \
-    sagemaker-containers==2.4.2 \
     $framework_support_installable \
     "sagemaker-tensorflow>=1.12,<1.13" \
     # Let's install TensorFlow separately in the end to avoid
@@ -129,5 +128,5 @@ RUN pip install --no-cache-dir -U \
 RUN ldconfig /usr/local/cuda-9.0/targets/x86_64-linux/lib/stubs && \
     HOROVOD_GPU_ALLREDUCE=NCCL  HOROVOD_WITH_TENSORFLOW=1 pip install --no-cache-dir horovod && \
     ldconfig
-    
-ENV SAGEMAKER_TRAINING_MODULE sagemaker_tensorflow_container.training:main
+
+ENV SAGEMAKER_TRAINING_MODULE sagemaker_tensorflow_container.training:main

setup.py

@@ -49,7 +49,7 @@ def read(fname):
         'Programming Language :: Python :: 3.6',
     ],
 
-    install_requires=['sagemaker-containers>=2.4.4', 'numpy', 'scipy', 'sklearn',
+    install_requires=['sagemaker-containers>=2.4.6', 'numpy', 'scipy', 'sklearn',
                       'pandas', 'Pillow', 'h5py'],
     extras_require={
         'test': ['tox', 'flake8', 'pytest', 'pytest-cov', 'pytest-xdist', 'mock',

test/integration/sagemaker/test_tuning_model_dir.py

@@ -0,0 +1,44 @@
+# Copyright 2019 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.
+from __future__ import absolute_import
+
+import os
+
+from sagemaker.tensorflow import TensorFlow
+from sagemaker.tuner import HyperparameterTuner, IntegerParameter
+
+
+def test_model_dir_with_training_job_name(sagemaker_session, ecr_image, instance_type, framework_version):
+    resource_path = os.path.join(os.path.dirname(__file__), '../..', 'resources')
+    script = os.path.join(resource_path, 'tuning_model_dir', 'entry.py')
+
+    estimator = TensorFlow(entry_point=script,
+                           role='SageMakerRole',
+                           train_instance_type=instance_type,
+                           train_instance_count=1,
+                           image_name=ecr_image,
+                           framework_version=framework_version,
+                           py_version='py3',
+                           sagemaker_session=sagemaker_session)
+
+    tuner = HyperparameterTuner(estimator=estimator,
+                                objective_metric_name='accuracy',
+                                hyperparameter_ranges={'arbitrary_value': IntegerParameter(0, 1)},
+                                metric_definitions=[{'Name': 'accuracy', 'Regex': 'accuracy=([01])'}],
+                                max_jobs=1,
+                                max_parallel_jobs=1,
+                                base_tuning_job_name='test-tf-tuning-model-dir')
+
+    # User script has logic to check for the correct model_dir
+    tuner.fit()
+    tuner.wait()

test/resources/tuning_model_dir/entry.py

@@ -0,0 +1,26 @@
+# Copyright 2019 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.
+from __future__ import absolute_import
+
+import argparse
+import os
+
+parser = argparse.ArgumentParser()
+parser.add_argument('--model_dir', type=str)
+parser.add_argument('--arbitrary_value', type=int, default=0)
+args = parser.parse_args()
+
+assert os.environ['TRAINING_JOB_NAME'] in args.model_dir, 'model_dir not unique to training job: %s' % args.model_dir
+
+# For the "hyperparameter tuning" to work
+print('accuracy=1')