Fetch resource variable fix (#276)

* Fixing SavedModelBundleTest.pythonTfFunction.

* Fixed 3 javadoc errors in tensorflow-core-api's hand written code.

* Changing the surefire parameters for tensorflow-core-api.

Author: Craigacp

tensorflow-core/tensorflow-core-api/pom.xml

@@ -388,8 +388,10 @@
           </execution>
         </executions>
         <configuration>
-          <!-- Activate the use of TCP to transmit events to the plugin -->
-          <forkNode implementation="org.apache.maven.plugin.surefire.extensions.SurefireForkNodeFactory"/>
+            <!-- Activate the use of TCP to transmit events to the plugin -->
+            <!-- disabled as it appears to cause intermittent test failures in GitHub Actions
+            <forkNode implementation="org.apache.maven.plugin.surefire.extensions.SurefireForkNodeFactory"/>
+            -->
           <additionalClasspathElements>
             <additionalClasspathElement>${project.build.directory}/${project.artifactId}-${project.version}-${native.classifier}.jar</additionalClasspathElement>
             <!-- Note: the following path is not accessible in deploying profile, so other libraries like

tensorflow-core/tensorflow-core-api/src/main/java/org/tensorflow/TensorMapper.java

@@ -26,7 +26,7 @@
  * <p>Usage of this class is reserved for internal purposes only.
  *
  * @param <T> tensor type mapped by this object
- * @see {@link TType}
+ * @see TType
  */
 public abstract class TensorMapper<T extends TType> {
 

tensorflow-core/tensorflow-core-api/src/main/java/org/tensorflow/op/core/Zeros.java

@@ -30,9 +30,9 @@
  * An operator creating a constant initialized with zeros of the shape given by `dims`.
  * 
  * <p>For example, the following expression
- * <pre>{@code tf.zeros(tf.constant(shape), TFloat32.class)</pre>
+ * <pre>{@code tf.zeros(tf.constant(shape), TFloat32.class)}</pre>
  * is the equivalent of
- * <pre>{@code tf.fill(tf.constant(shape), tf.constant(0.0f))</pre>
+ * <pre>{@code tf.fill(tf.constant(shape), tf.constant(0.0f))}</pre>
  *
  * @param <T> constant type
  */

tensorflow-core/tensorflow-core-api/src/main/java/org/tensorflow/types/family/TType.java

@@ -28,7 +28,7 @@
  *
  * <p>Subinterfaces of {@code TType} are propagated as a generic parameter to various entities of
  * TensorFlow to identify the type of the tensor they carry. For example, a
- * {@link org.tensorflow.Operand Operand<TFloat32>} is an operand which outputs a 32-bit floating
+ * {@link org.tensorflow.Operand Operand&lt;TFloat32&gt;} is an operand which outputs a 32-bit floating
  * point tensor. This parameter ensure type-compatibility between operands of a computation at
  * compile-time. For example:
  *

tensorflow-core/tensorflow-core-api/src/test/java/org/tensorflow/SavedModelBundleTest.java

@@ -290,7 +290,7 @@ public void pythonTfFunction() {
        *   Signature name used for saving 'add', argument names 'a' and 'b'
        */
       ConcreteFunction add = bundle.function("add");
-      Map<String, Tensor> args = new HashMap();
+      Map<String, Tensor> args = new HashMap<>();
       try (TFloat32 a = TFloat32.scalarOf(10.0f);
           TFloat32 b = TFloat32.scalarOf(15.5f)) {
         args.put("a", a);
@@ -301,14 +301,19 @@ public void pythonTfFunction() {
           assertEquals(25.5f, c.getFloat());
         }
       }
+      args.clear();
 
       // variable unwrapping happens in Session, which is used by ConcreteFunction.call
       ConcreteFunction getVariable = bundle.function("get_variable");
-      try (TFloat32 v = (TFloat32) getVariable.call(new HashMap<>())
-          .get(getVariable.signature().outputNames().iterator().next())) {
-        assertEquals(2f, v.getFloat());
+      try (TFloat32 dummy = TFloat32.scalarOf(1.0f)) {
+        args.put("dummy",dummy);
+        // TF functions always require an input, so we supply a dummy one here
+        // This test actually checks that resource variables can be loaded correctly.
+        try (TFloat32 v = (TFloat32) getVariable.call(args)
+                    .get(getVariable.signature().outputNames().iterator().next())) {
+          assertEquals(2f, v.getFloat());
+        }
       }
-
     }
   }
 

tensorflow-core/tensorflow-core-api/src/test/resources/saved_model_using_python/model/saved_model.pb

@@ -22,6 +22,7 @@
 
 import tensorflow as tf
 
+
 class MyModel(tf.keras.Model):
   def __init__(self):
     super(MyModel, self).__init__()
@@ -42,8 +43,7 @@ def get_scalar(self, x):
   def get_vector(self, x):
     return self.const_vector + x
 
-  @tf.function(input_signature=[
-    tf.TensorSpec(shape=None, dtype=tf.float32, name='input')])
+  @tf.function(input_signature=[tf.TensorSpec(shape=None, dtype=tf.float32, name='input')])
   def get_matrix(self, x):
     return self.const_matrix + x
 
@@ -53,10 +53,14 @@ def get_matrix(self, x):
   def add(self, a, b):
     return a + b
 
-  @tf.function(input_signature=[])
-  def get_variable(self):
+  #TF functions always require an input
+  @tf.function(input_signature=[
+    tf.TensorSpec(shape=None, dtype=tf.float32, name='dummy')
+  ])
+  def get_variable(self, dummy):
     return self.variable
 
+
 model = MyModel()
 
 signatures = {