Add integ tests for tuning jobs (aws#220)

Author: ChoiByungWook

CHANGELOG.rst

@@ -6,6 +6,7 @@ CHANGELOG
 ========
 
 * bug-fix: Unit Tests: Improve unit test runtime
+* bug-fix: Estimators: Fix attach for LDA
 
 1.4.1
 =====

src/sagemaker/amazon/lda.py

@@ -78,8 +78,10 @@ def __init__(self, role, train_instance_type, num_topics,
             tol (float): Optional. Target error tolerance for the ALS phase of the algorithm.
             **kwargs: base class keyword argument values.
         """
-
         # this algorithm only supports single instance training
+        if kwargs.pop('train_instance_count', 1) != 1:
+            print('LDA only supports single instance training. Defaulting to 1 {}.'.format(train_instance_type))
+
         super(LDA, self).__init__(role, 1, train_instance_type, **kwargs)
         self.num_topics = num_topics
         self.alpha0 = alpha0

tests/conftest.py

@@ -16,6 +16,7 @@
 
 import boto3
 import pytest
+from botocore.config import Config
 
 from sagemaker import Session
 from sagemaker.local import LocalSession
@@ -32,7 +33,7 @@ def pytest_addoption(parser):
 @pytest.fixture(scope='session')
 def sagemaker_client_config(request):
     config = request.config.getoption('--sagemaker-client-config')
-    return json.loads(config) if config else None
+    return json.loads(config) if config else dict()
 
 
 @pytest.fixture(scope='session')
@@ -50,6 +51,7 @@ def boto_config(request):
 @pytest.fixture(scope='session')
 def sagemaker_session(sagemaker_client_config, sagemaker_runtime_config, boto_config):
     boto_session = boto3.Session(**boto_config) if boto_config else boto3.Session(region_name=DEFAULT_REGION)
+    sagemaker_client_config.setdefault('config', Config(retries=dict(max_attempts=10)))
     sagemaker_client = boto_session.client('sagemaker', **sagemaker_client_config) if sagemaker_client_config else None
     runtime_client = (boto_session.client('sagemaker-runtime', **sagemaker_runtime_config) if sagemaker_runtime_config
                       else None)

tests/data/chainer_mnist/mnist.py

@@ -72,6 +72,7 @@ def _preprocess_mnist(raw, withlabel, ndim, scale, image_dtype, label_dtype, rgb
     parser.add_argument('--epochs', type=int, default=20)
     parser.add_argument('--frequency', type=int, default=20)
     parser.add_argument('--batch-size', type=int, default=100)
+    parser.add_argument('--alpha', type=float, default=0.001)
     parser.add_argument('--model-dir', type=str, default=env.model_dir)
 
     parser.add_argument('--train', type=str, default=env.channel_input_dirs['train'])
@@ -103,7 +104,7 @@ def _preprocess_mnist(raw, withlabel, ndim, scale, image_dtype, label_dtype, rgb
         chainer.cuda.get_device_from_id(0).use()
 
     # Setup an optimizer
-    optimizer = chainer.optimizers.Adam()
+    optimizer = chainer.optimizers.Adam(alpha=args.alpha)
     optimizer.setup(model)
 
     # Load the MNIST dataset

tests/data/iris/iris-dnn-classifier.py

@@ -18,10 +18,12 @@
 
 
 def estimator_fn(run_config, hyperparameters):
-    input_tensor_name = hyperparameters['input_tensor_name']
+    input_tensor_name = hyperparameters.get('input_tensor_name', 'inputs')
+    learning_rate = hyperparameters.get('learning_rate', 0.05)
     feature_columns = [tf.feature_column.numeric_column(input_tensor_name, shape=[4])]
     return tf.estimator.DNNClassifier(feature_columns=feature_columns,
                                       hidden_units=[10, 20, 10],
+                                      optimizer=tf.train.AdagradOptimizer(learning_rate=learning_rate),
                                       n_classes=3,
                                       config=run_config)