Merge pull request #548 from kirilg/branch_164134747

Upstream changes

Author: kirilg

tensorflow_serving/example/BUILD

@@ -26,21 +26,6 @@ py_library(
     srcs_version = "PY2AND3",
 )
 
-# TODO(b/32628014): remove mnist_export after we no longer support
-# SessionBundle.
-py_binary(
-    name = "mnist_export",
-    srcs = [
-        "mnist_export.py",
-    ],
-    srcs_version = "PY2AND3",
-    deps = [
-        ":mnist_input_data",
-        "@org_tensorflow//tensorflow:tensorflow_py",
-        "@org_tensorflow//tensorflow/contrib/session_bundle:exporter",
-    ],
-)
-
 py_binary(
     name = "mnist_saved_model",
     srcs = [
@@ -67,21 +52,6 @@ py_binary(
     ],
 )
 
-# TODO(b/32628014): remove inception_export after we no longer support
-# SessionBundle.
-py_binary(
-    name = "inception_export",
-    srcs = [
-        "inception_export.py",
-    ],
-    srcs_version = "PY2AND3",
-    deps = [
-        "@inception_model//inception",
-        "@org_tensorflow//tensorflow:tensorflow_py",
-        "@org_tensorflow//tensorflow/contrib/session_bundle:exporter",
-    ],
-)
-
 py_binary(
     name = "inception_saved_model",
     srcs = [

tensorflow_serving/example/inception_export.py

@@ -14,14 +14,15 @@
 # ==============================================================================
 
 #!/usr/bin/env python2.7
-"""Train and export a simple Softmax Regression TensorFlow model.
+r"""Train and export a simple Softmax Regression TensorFlow model.
 
 The model is from the TensorFlow "MNIST For ML Beginner" tutorial. This program
 simply follows all its training instructions, and uses TensorFlow SavedModel to
 export the trained model with proper signatures that can be loaded by standard
 tensorflow_model_server.
 
-Usage: mnist_export.py [--training_iteration=x] [--model_version=y] export_dir
+Usage: mnist_saved_model.py [--training_iteration=x] [--model_version=y] \
+    export_dir
 """
 
 import os

tensorflow_serving/example/mnist_export.py

@@ -308,7 +308,7 @@ int main(int argc, char** argv) {
   tensorflow::string model_name = "default";
   tensorflow::int32 file_system_poll_wait_seconds = 1;
   tensorflow::string model_base_path;
-  bool use_saved_model = true;
+  const bool use_saved_model = true;
   // Tensorflow session parallelism of zero means that both inter and intra op
   // thread pools will be auto configured.
   tensorflow::int64 tensorflow_session_parallelism = 0;
@@ -338,11 +338,6 @@ int main(int argc, char** argv) {
                        &file_system_poll_wait_seconds,
                        "interval in seconds between each poll of the file "
                        "system for new model version"),
-      tensorflow::Flag("use_saved_model", &use_saved_model,
-                       "If true, use SavedModel in the server; otherwise, use "
-                       "SessionBundle. It is used by tensorflow serving team "
-                       "to control the rollout of SavedModel and is not "
-                       "expected to be set by users directly."),
       tensorflow::Flag("tensorflow_session_parallelism",
                        &tensorflow_session_parallelism,
                        "Number of threads to use for running a "
@@ -353,8 +348,7 @@ int main(int argc, char** argv) {
                        "If non-empty, read an ascii PlatformConfigMap protobuf "
                        "from the supplied file name, and use that platform "
                        "config instead of the Tensorflow platform. (If used, "
-                       "--enable_batching and --use_saved_model are "
-                       "ignored.)")};
+                       "--enable_batching is ignored.)")};
   string usage = tensorflow::Flags::Usage(argv[0], flag_list);
   const bool parse_result = tensorflow::Flags::Parse(&argc, argv, flag_list);
   if (!parse_result || (model_base_path.empty() && model_config_file.empty())) {