Updated example to support latest TF framework

Author: FarhanTejani

sagemaker-python-sdk/tensorflow_resnet_cifar10_with_tensorboard/source_dir/resnet_cifar_10.py

@@ -133,13 +133,13 @@ def _input_from_files(mode, batch_size, data_dir):
     if mode == tf.estimator.ModeKeys.TRAIN:
         dataset = dataset.repeat()
 
-    dataset = dataset.map(_dataset_parser, num_threads=1,
-                          output_buffer_size=2 * batch_size)
+    dataset = dataset.map(_dataset_parser)
+    dataset.prefetch(2 * batch_size)
 
     # For training, preprocess the image and shuffle.
     if mode == tf.estimator.ModeKeys.TRAIN:
-        dataset = dataset.map(_train_preprocess_fn, num_threads=1,
-                              output_buffer_size=2 * batch_size)
+        dataset = dataset.map(_train_preprocess_fn)
+        dataset.prefetch(2 * batch_size)
 
         # Ensure that the capacity is sufficiently large to provide good random
         # shuffling.
@@ -148,9 +148,8 @@ def _input_from_files(mode, batch_size, data_dir):
 
     # Subtract off the mean and divide by the variance of the pixels.
     dataset = dataset.map(
-        lambda image, label: (tf.image.per_image_standardization(image), label),
-        num_threads=1,
-        output_buffer_size=2 * batch_size)
+        lambda image, label: (tf.image.per_image_standardization(image), label))
+    dataset.prefetch(2 * batch_size)
 
     # Batch results by up to batch_size, and then fetch the tuple from the
     # iterator.
@@ -203,7 +202,7 @@ def _dataset_parser(value):
 def _record_dataset(filenames):
     """Returns an input pipeline Dataset from `filenames`."""
     record_bytes = HEIGHT * WIDTH * DEPTH + 1
-    return tf.contrib.data.FixedLengthRecordDataset(filenames, record_bytes)
+    return tf.data.FixedLengthRecordDataset(filenames, record_bytes)
 
 
 def _filenames(mode, data_dir):

sagemaker-python-sdk/tensorflow_resnet_cifar10_with_tensorboard/tensorflow_resnet_cifar10_with_tensorboard.ipynb

@@ -118,7 +118,6 @@
     "estimator = TensorFlow(entry_point='resnet_cifar_10.py',\n",
     "                       source_dir=source_dir,\n",
     "                       role=role,\n",
-    "                       framework_version='1.6',\n",
     "                       hyperparameters={'throttle_secs': 30},\n",
     "                       training_steps=1000, evaluation_steps=100,\n",
     "                       train_instance_count=2, train_instance_type='ml.c4.xlarge', \n",
@@ -146,9 +145,7 @@
   },
   {
    "cell_type": "markdown",
-   "metadata": {
-    "collapsed": true
-   },
+   "metadata": {},
    "source": [
     "# Deploy the trained model to prepare for predictions\n",
     "\n",
@@ -225,7 +222,7 @@
    "name": "python",
    "nbconvert_exporter": "python",
    "pygments_lexer": "ipython2",
-   "version": "2.7.11"
+   "version": "2.7.14"
   },
   "notice": "Copyright 2017 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."
  },