Fix book.yaml

Remove 'via'

PiperOrigin-RevId: 220486241

Author: lamberta

tensorflow_serving/g3doc/_book.yaml

@@ -25,12 +25,8 @@ upper_tabs:
         path: /serving/setup
       - title: Serve a TensorFlow model
         path: /serving/serving_basic
-      - title: REST API
-        path: /serving/api_rest
       - title: Build a TensorFlow ModelServer
         path: /serving/serving_advanced
-      - title: Use TensorFlow Serving with Docker
-        path: /serving/docker
       - title: Use TensorFlow Serving with Kubernetes
         path: /serving/serving_kubernetes
       - title: Create a new kind of servable

tensorflow_serving/g3doc/custom_servable.md

@@ -30,7 +30,7 @@ document).
 
 In addition to your `Loader`, you will need to define a `SourceAdapter` that
 instantiates a `Loader` from a given storage path. Most simple use-cases can
-specify the two objects concisely via the `SimpleLoaderSourceAdapter` class
+specify the two objects concisely with the `SimpleLoaderSourceAdapter` class
 (in `core/simple_loader.h`). Advanced use-cases may opt to specify `Loader` and
 `SourceAdapter` classes separately using the lower-level APIs, e.g. if the
 `SourceAdapter` needs to retain some state, and/or if state needs to be shared

tensorflow_serving/g3doc/docker.md

@@ -1,6 +1,6 @@
-# Using TensorFlow Serving via Docker
+# Using TensorFlow Serving with Docker
 
-One of the easiest ways to get started using TensorFlow Serving is via
+One of the easiest ways to get started using TensorFlow Serving is with
 [Docker](http://www.docker.com/).
 
 ## Installing Docker
@@ -136,8 +136,8 @@ deploy and will load your model for serving on startup.
 
 ### Serving example
 
-Let's run through a full example where we load a SavedModel and call it via the
-REST API. First pull the serving image:
+Let's run through a full example where we load a SavedModel and call it using
+the REST API. First pull the serving image:
 
 ```shell
 docker pull tensorflow/serving
@@ -209,7 +209,7 @@ details, see [running a serving image](#running-a-serving-image).
 ### GPU Serving example
 
 Let's run through a full example where we load a model with GPU-bound ops and
-call it via the REST API.
+call it using the REST API.
 
 First install [`nvidia-docker`](#install-nvidia-docker). Next you can pull the
 latest TensorFlow Serving GPU docker image by running:

tensorflow_serving/g3doc/overview.md

@@ -121,7 +121,7 @@ TensorFlow Serving Managers provide a simple, narrow interface --
 
 ### Core
 
-**TensorFlow Serving Core** manages (via standard TensorFlow Serving APIs) the
+Using the standard TensorFlow Serving APis, *TensorFlow Serving Core* manages the
 following aspects of servables:
 
   * lifecycle

tensorflow_serving/g3doc/setup.md

@@ -4,7 +4,7 @@
 
 ### Installing using Docker
 
-The easiest and most straight-forward way of using TensorFlow Serving is via
+The easiest and most straight-forward way of using TensorFlow Serving is with
 [Docker images](docker.md). We highly recommend this route unless you have
 specific needs that are not addressed by running in a container.