From 28a34dd19fd88c892e7b3ea62489bb062fb9b11b Mon Sep 17 00:00:00 2001
From: Jim Clarke <JimClarke5@me.com>
Date: Wed, 3 Mar 2021 11:21:50 -0500
Subject: [PATCH 01/13] Clean up generics, remove generics from class and fix
 call method to have generic.

---
 .../annotations/org/tensorflow/op/Ops.java    | 12 ++--
 .../framework/activations/Activation.java     | 26 ++++-----
 .../tensorflow/framework/activations/ELU.java | 12 ++--
 .../framework/activations/Exponential.java    |  9 ++-
 .../framework/activations/HardSigmoid.java    |  9 ++-
 .../framework/activations/Linear.java         |  8 +--
 .../framework/activations/ReLU.java           |  6 +-
 .../framework/activations/SELU.java           |  8 +--
 .../framework/activations/Sigmoid.java        |  9 ++-
 .../framework/activations/Softmax.java        |  9 ++-
 .../framework/activations/Softplus.java       |  6 +-
 .../framework/activations/Softsign.java       |  9 ++-
 .../framework/activations/Swish.java          |  7 +--
 .../framework/activations/Tanh.java           |  8 +--
 .../initializers/BaseInitializer.java         |  3 +-
 .../framework/initializers/Constant.java      |  9 ++-
 .../framework/initializers/Glorot.java        |  8 +--
 .../tensorflow/framework/initializers/He.java | 12 ++--
 .../framework/initializers/Identity.java      |  8 +--
 .../framework/initializers/Initializer.java   | 11 ++--
 .../framework/initializers/LeCun.java         | 11 ++--
 .../framework/initializers/Ones.java          |  9 ++-
 .../framework/initializers/Orthogonal.java    | 38 +++++++-----
 .../framework/initializers/RandomNormal.java  | 29 ++++++----
 .../framework/initializers/RandomUniform.java | 39 ++++++++-----
 .../initializers/TruncatedNormal.java         | 33 +++++++----
 .../initializers/VarianceScaling.java         | 38 +++++++-----
 .../framework/initializers/Zeros.java         |  8 +--
 .../framework/activations/ELUTest.java        | 10 +---
 .../activations/ExponentialTest.java          |  8 +--
 .../activations/HardSigmoidTest.java          |  8 +--
 .../framework/activations/LinearTest.java     |  6 +-
 .../framework/activations/ReLUTest.java       | 20 +++----
 .../framework/activations/SELUTest.java       |  8 +--
 .../framework/activations/SigmoidTest.java    |  7 +--
 .../framework/activations/SoftmaxTest.java    | 21 +++----
 .../framework/activations/SoftplusTest.java   |  4 +-
 .../framework/activations/SoftsignTest.java   |  4 +-
 .../framework/activations/SwishTest.java      |  8 +--
 .../framework/activations/TanhTest.java       |  4 +-
 .../framework/initializers/ConstantTest.java  | 16 ++---
 .../framework/initializers/GlorotTest.java    | 15 +++--
 .../framework/initializers/HeTest.java        | 15 +++--
 .../framework/initializers/IdentityTest.java  | 10 +---
 .../framework/initializers/LeCunTest.java     | 14 ++---
 .../framework/initializers/OnesTest.java      | 16 ++---
 .../initializers/OrthogonalTest.java          | 10 +---
 .../initializers/RandomNormalTest.java        |  9 +--
 .../initializers/RandomUniformTest.java       | 12 ++--
 .../initializers/TruncatedNormalTest.java     |  9 +--
 .../initializers/VarianceScalingTest.java     | 58 ++++++++-----------
 .../framework/initializers/ZerosTest.java     | 16 ++---
 52 files changed, 327 insertions(+), 375 deletions(-)

diff --git a/tensorflow-core/tensorflow-core-api/src/gen/annotations/org/tensorflow/op/Ops.java b/tensorflow-core/tensorflow-core-api/src/gen/annotations/org/tensorflow/op/Ops.java
index ea3ef31313e..acbae4dac6b 100644
--- a/tensorflow-core/tensorflow-core-api/src/gen/annotations/org/tensorflow/op/Ops.java
+++ b/tensorflow-core/tensorflow-core-api/src/gen/annotations/org/tensorflow/op/Ops.java
@@ -354,20 +354,20 @@ public final class Ops {
 
   public final SparseOps sparse;
 
-  public final TpuOps tpu;
-
   public final BitwiseOps bitwise;
 
+  public final TpuOps tpu;
+
   public final MathOps math;
 
   public final AudioOps audio;
 
   public final SignalOps signal;
 
-  public final TrainOps train;
-
   public final QuantizationOps quantization;
 
+  public final TrainOps train;
+
   private final Scope scope;
 
   private Ops(Scope scope) {
@@ -385,13 +385,13 @@ private Ops(Scope scope) {
     random = new RandomOps(this);
     strings = new StringsOps(this);
     sparse = new SparseOps(this);
-    tpu = new TpuOps(this);
     bitwise = new BitwiseOps(this);
+    tpu = new TpuOps(this);
     math = new MathOps(this);
     audio = new AudioOps(this);
     signal = new SignalOps(this);
-    train = new TrainOps(this);
     quantization = new QuantizationOps(this);
+    train = new TrainOps(this);
   }
 
   /**
diff --git a/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Activation.java b/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Activation.java
index e1482a51a8a..104e3726ee4 100644
--- a/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Activation.java
+++ b/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Activation.java
@@ -20,13 +20,8 @@
 
 /**
  * Abstract base class for Activations
- *
- * <p><b>Note:</b> The {@link #tf} attribute must be set prior to invoking the call method. See
- * {@link #setTF(Ops)} and the constructor {@link #Activation(Ops)}.
- *
- * @param <T> the data type of the activation
  */
-public abstract class Activation<T extends TNumber> {
+public abstract class Activation {
 
   /** The TensorFlow Ops */
   protected Ops tf;
@@ -41,21 +36,21 @@ protected Activation(Ops tf) {
   }
 
   /**
-   * Sets the TensorFlow Ops
+   * Gets the TensorFlow Ops
    *
-   * @param tf the TensorFlow Ops
+   * @return the TensorFlow Ops
    */
-  protected void setTF(Ops tf) {
-    this.tf = tf;
+  protected Ops getTF() {
+    return this.tf;
   }
 
   /**
-   * Gets the TensorFlow Ops
+   * Sets the TensorFlow Ops
    *
-   * @return the TensorFlow Ops
+   * @param tf the TensorFlow Ops
    */
-  protected Ops getTF() {
-    return this.tf;
+  protected void setTF(Ops tf) {
+    this.tf = tf;
   }
 
   /**
@@ -63,6 +58,7 @@ protected Ops getTF() {
    *
    * @param input the input tensor
    * @return The operand for the activation
+   * @param <T> the data type of the input and result
    */
-  public abstract Operand<T> call(Operand<T> input);
+  public abstract <T extends TNumber> Operand<T> call(Operand<T> input);
 }
diff --git a/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/ELU.java b/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/ELU.java
index 2f2f16f2752..e7ad39fde62 100644
--- a/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/ELU.java
+++ b/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/ELU.java
@@ -17,7 +17,7 @@
 import org.tensorflow.Operand;
 import org.tensorflow.op.Ops;
 import org.tensorflow.types.TBool;
-import org.tensorflow.types.family.TFloating;
+import org.tensorflow.types.family.TNumber;
 
 /**
  * Exponential linear unit.
@@ -44,11 +44,10 @@
  *     Operand&lt;TFloat32&gt; result = elu.call(input);
  * </pre>
  *
- * @param <T> the data type of the activation
  * @see <a href="https://arxiv.org/abs/1511.07289">Clevert et al, 2016, Fast and Accurate Deep
  *     Network Learning by Exponential Linear Units (ELUs)</a>
  */
-public class ELU<T extends TFloating> extends Activation<T> {
+public class ELU extends Activation {
 
   private static final double ALPHA_DEFAULT = 1.0;
 
@@ -83,11 +82,12 @@ public ELU(Ops tf, double alpha) {
    * @return The operand for the activation
    */
   @Override
-  public Operand<T> call(Operand<T> input) {
+  public <T extends TNumber> Operand<T> call(Operand<T> input) {
 
     Operand<T> result = tf.nn.elu(input);
-    if (alpha == 1.0) return result;
-    else {
+    if (alpha == 1.0) {
+      return result;
+    } else {
       Class<T> inputType = input.type();
       Operand<T> y = tf.math.mul(result, tf.dtypes.cast(tf.constant(alpha), inputType));
       Operand<TBool> cond = tf.math.greater(result, tf.dtypes.cast(tf.constant(0), inputType));
diff --git a/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Exponential.java b/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Exponential.java
index d5fdff36c61..2debca9ddac 100644
--- a/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Exponential.java
+++ b/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Exponential.java
@@ -16,7 +16,7 @@
 
 import org.tensorflow.Operand;
 import org.tensorflow.op.Ops;
-import org.tensorflow.types.family.TFloating;
+import org.tensorflow.types.family.TNumber;
 
 /**
  * Exponential activation function.
@@ -30,10 +30,8 @@
  *   Operand&lt;TFloat32&gt; result = exp.call(input);
  *   // result is [0.04978707f,  0.36787945f,  1.f,  2.7182817f, 20.085537f]
  * </pre>
- *
- * @param <T> the data type of the activation
  */
-public class Exponential<T extends TFloating> extends Activation<T> {
+public class Exponential extends Activation {
 
   /**
    * Creates an Exponential activation.
@@ -49,9 +47,10 @@ public Exponential(Ops tf) {
    *
    * @param input the input tensor
    * @return an Operand for the exponential activation: <code>exp(x)</code>.
+   * @param <T> the data type of the activation
    */
   @Override
-  public Operand<T> call(Operand<T> input) {
+  public <T extends TNumber> Operand<T> call(Operand<T> input) {
     return tf.math.exp(input);
   }
 }
diff --git a/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/HardSigmoid.java b/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/HardSigmoid.java
index 0b7cf573b8e..0e2fd6f9342 100644
--- a/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/HardSigmoid.java
+++ b/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/HardSigmoid.java
@@ -16,7 +16,7 @@
 
 import org.tensorflow.Operand;
 import org.tensorflow.op.Ops;
-import org.tensorflow.types.family.TFloating;
+import org.tensorflow.types.family.TNumber;
 
 /**
  * Hard sigmoid activation.
@@ -40,10 +40,8 @@
  *     Operand&lt;TFloat32&gt; result = hardSigmoid.call(input);
  *     // result is [0.f , 0.3f, 0.5f, 0.7f, 1.f]
  * </pre>
- *
- * @param <T> the data type of the result
  */
-public class HardSigmoid<T extends TFloating> extends Activation<T> {
+public class HardSigmoid extends Activation {
 
   /**
    * Creates Hard sigmoid activation.
@@ -59,9 +57,10 @@ public HardSigmoid(Ops tf) {
    *
    * @param input the input tensor
    * @return The operand for the activation
+   * @param <T> the data type of the result
    */
   @Override
-  public Operand<T> call(Operand<T> input) {
+  public <T extends TNumber> Operand<T> call(Operand<T> input) {
     Class<T> inputType = input.type();
     Operand<T> point2 = tf.dtypes.cast(tf.constant(0.2), inputType);
     Operand<T> point5 = tf.dtypes.cast(tf.constant(0.5), inputType);
diff --git a/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Linear.java b/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Linear.java
index d907397995d..06aba774423 100644
--- a/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Linear.java
+++ b/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Linear.java
@@ -19,9 +19,9 @@
 import org.tensorflow.types.family.TNumber;
 
 /**
- * Linear activation function  (pass-through).
+ * Linear activation function (pass-through).
  *
- * <p>The linear activation returns its input. It is also known as the Identity activation function.</p>
+ * <p>The linear activation returns its input. It is also known as the Identity activation function.
  *
  * <p>For example:
  *
@@ -33,7 +33,7 @@
  *    // result is [-3.0f,-1.0f, 0.0f,1.0f,3.0f]
  * </pre>
  */
-public class Linear<U extends TNumber> extends Activation<U> {
+public class Linear extends Activation {
 
   /**
    * Creates a linear activation.
@@ -46,7 +46,7 @@ public Linear(Ops tf) {
 
   /** {@inheritDoc} */
   @Override
-  public Operand<U> call(Operand<U> input) {
+  public <T extends TNumber> Operand<T> call(Operand<T> input) {
     return input;
   }
 }
diff --git a/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/ReLU.java b/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/ReLU.java
index aef6ebf2992..974d266143f 100644
--- a/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/ReLU.java
+++ b/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/ReLU.java
@@ -55,10 +55,8 @@
  *     result = relu.call(input);
  *     // result is [-0.f, -0.f,  0.f,  0.f, 10.f]
  * </pre>
- *
- * @param <T> the data type of the result
  */
-public class ReLU<T extends TNumber> extends Activation<T> {
+public class ReLU extends Activation {
 
   public static final float ALPHA_DEFAULT = 0.0f;
   public static final float MAX_VALUE_DEFAULT = Float.NaN;
@@ -96,7 +94,7 @@ public ReLU(Ops tf, float alpha, float maxValue, float threshold) {
 
   /** {@inheritDoc} */
   @Override
-  public Operand<T> call(Operand<T> input) {
+  public <T extends TNumber> Operand<T> call(Operand<T> input) {
     Class<T> inputType = input.type();
 
     boolean clipMax = !Float.isNaN(maxValue);
diff --git a/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/SELU.java b/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/SELU.java
index f24731049fb..d3328aeb110 100644
--- a/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/SELU.java
+++ b/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/SELU.java
@@ -16,7 +16,7 @@
 
 import org.tensorflow.Operand;
 import org.tensorflow.op.Ops;
-import org.tensorflow.types.family.TFloating;
+import org.tensorflow.types.family.TNumber;
 
 /**
  * Scaled Exponential Linear Unit (SELU).
@@ -42,10 +42,9 @@
  * <p><b>Notes: </b> To be used together with the {@link
  * org.tensorflow.framework.initializers.LeCun} initializer with Normal Distribution.
  *
- * @param <T> the data type of the activation
  * @see <a href="https://arxiv.org/abs/1706.02515">Klambauer et al., 2017</a>
  */
-public class SELU<T extends TFloating> extends Activation<T> {
+public class SELU extends Activation {
 
   /**
    * Creates a Scaled Exponential Linear Unit (SELU) activation.
@@ -61,9 +60,10 @@ public SELU(Ops tf) {
    *
    * @param input the input tensor
    * @return The operand for the activation
+   * @param <T> the data type of the activation
    */
   @Override
-  public Operand<T> call(Operand<T> input) {
+  public <T extends TNumber> Operand<T> call(Operand<T> input) {
     return tf.nn.selu(input);
   }
 }
diff --git a/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Sigmoid.java b/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Sigmoid.java
index 5d507b38483..c4781ba50d9 100644
--- a/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Sigmoid.java
+++ b/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Sigmoid.java
@@ -16,7 +16,7 @@
 
 import org.tensorflow.Operand;
 import org.tensorflow.op.Ops;
-import org.tensorflow.types.family.TFloating;
+import org.tensorflow.types.family.TNumber;
 
 /**
  * Sigmoid activation. <code>sigmoid(x) = 1 / (1 + exp(-x))</code>.
@@ -38,10 +38,8 @@
  *     // result is [2.0611537e-09f, 2.6894143e-01f,
  *     //                 5.0000000e-01f,7.3105860e-01f, 1.f]
  * </pre>
- *
- * @param <T> the data type of the activation
  */
-public class Sigmoid<T extends TFloating> extends Activation<T> {
+public class Sigmoid extends Activation {
 
   /**
    * Creates a Sigmoid activation.
@@ -57,9 +55,10 @@ public Sigmoid(Ops tf) {
    *
    * @param input the input tensor
    * @return The operand for the activation
+   * @param <T> the data type of the activation
    */
   @Override
-  public Operand<T> call(Operand<T> input) {
+  public <T extends TNumber> Operand<T> call(Operand<T> input) {
     return tf.math.sigmoid(input);
   }
 }
diff --git a/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Softmax.java b/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Softmax.java
index 154e1ecc84a..8051cee9c04 100644
--- a/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Softmax.java
+++ b/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Softmax.java
@@ -19,7 +19,7 @@
 import org.tensorflow.op.Ops;
 import org.tensorflow.op.core.ReduceMax;
 import org.tensorflow.op.core.ReduceSum;
-import org.tensorflow.types.family.TFloating;
+import org.tensorflow.types.family.TNumber;
 
 /**
  * Softmax converts a real vector to a vector of categorical probabilities.
@@ -35,10 +35,8 @@
  * <p>The softmax of each vector x is computed as: <code>exp(x) / tf.sum(exp(x))</code>.
  *
  * <p>The input values in are the log-odds of the resulting probability.
- *
- * @param <T> the data type of the activation
  */
-public class Softmax<T extends TFloating> extends Activation<T> {
+public class Softmax extends Activation {
 
   private static final int AXIS_DEFAULT = -1;
 
@@ -70,9 +68,10 @@ public Softmax(Ops tf, int axis) {
    *
    * @param input the input tensor
    * @return The operand for the activation
+   * @param <T> the data type of the activation
    */
   @Override
-  public Operand<T> call(Operand<T> input) {
+  public <T extends TNumber> Operand<T> call(Operand<T> input) {
     Shape shape = input.shape();
     int numDimensions = shape.numDimensions();
     if (numDimensions == 2) {
diff --git a/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Softplus.java b/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Softplus.java
index 65a183ea047..762c0ae6bcd 100644
--- a/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Softplus.java
+++ b/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Softplus.java
@@ -16,7 +16,7 @@
 
 import org.tensorflow.Operand;
 import org.tensorflow.op.Ops;
-import org.tensorflow.types.family.TFloating;
+import org.tensorflow.types.family.TNumber;
 
 /**
  * Softplus activation function, <code>softplus(x) = log(exp(x) + 1)</code>.
@@ -32,7 +32,7 @@
  *     //                 1.3132616e+00f, 2.0000000e+01f]
  * </pre>
  */
-public class Softplus<T extends TFloating> extends Activation<T> {
+public class Softplus extends Activation {
 
   /**
    * Creates a Softplus activation function.
@@ -50,7 +50,7 @@ public Softplus(Ops tf) {
    * @return The operand for the activation
    */
   @Override
-  public Operand<T> call(Operand<T> input) {
+  public <T extends TNumber> Operand<T> call(Operand<T> input) {
     return tf.math.softplus(input);
   }
 }
diff --git a/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Softsign.java b/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Softsign.java
index 1f691e71862..b6e9f874914 100644
--- a/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Softsign.java
+++ b/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Softsign.java
@@ -16,7 +16,7 @@
 
 import org.tensorflow.Operand;
 import org.tensorflow.op.Ops;
-import org.tensorflow.types.family.TFloating;
+import org.tensorflow.types.family.TNumber;
 
 /**
  * Softsign activation function, <code>softsign(x) = x / (abs(x) + 1)</code>.
@@ -30,10 +30,8 @@
  *     Operand&lt;TFloat32&gt; result = softsign.call(input);
  *     // result is [-0.5f, 0.f, 0.5f]
  * </pre>
- *
- * @param <T> the data type of the activation
  */
-public class Softsign<T extends TFloating> extends Activation<T> {
+public class Softsign extends Activation {
 
   /**
    * Creates a Softsign activation.
@@ -49,9 +47,10 @@ public Softsign(Ops tf) {
    *
    * @param input the input tensor
    * @return The operand for the activation
+   * @param <T> the data type of the activation
    */
   @Override
-  public Operand<T> call(Operand<T> input) {
+  public <T extends TNumber> Operand<T> call(Operand<T> input) {
     return tf.nn.softsign(input);
   }
 }
diff --git a/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Swish.java b/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Swish.java
index d9f73a422d5..5a6fc4c7765 100644
--- a/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Swish.java
+++ b/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Swish.java
@@ -16,7 +16,7 @@
 
 import org.tensorflow.Operand;
 import org.tensorflow.op.Ops;
-import org.tensorflow.types.family.TFloating;
+import org.tensorflow.types.family.TNumber;
 
 /**
  * Swish activation function. <code>swish(x) = x * sigmoid(x)</code>.
@@ -37,10 +37,9 @@
  *
  * </pre>
  *
- * @param <T> the data type of the activation
  * @see <a href="https://arxiv.org/abs/1710.05941">Ramachandran et al., 2017</a>
  */
-public class Swish<T extends TFloating> extends Activation<T> {
+public class Swish extends Activation {
 
   /**
    * Creates a Swish activation, <code>swish(x) = x * sigmoid(x)</code>.
@@ -57,7 +56,7 @@ public Swish(Ops tf) {
 
   /** {@inheritDoc} */
   @Override
-  public Operand<T> call(Operand<T> input) {
+  public <T extends TNumber> Operand<T> call(Operand<T> input) {
 
     // TODO Python Keras returns a "grad", which is an optimization not implemented in Java.
     return tf.math.mul(input, tf.math.sigmoid(input));
diff --git a/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Tanh.java b/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Tanh.java
index 4fe02eed048..a485638aa4a 100644
--- a/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Tanh.java
+++ b/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Tanh.java
@@ -16,7 +16,7 @@
 
 import org.tensorflow.Operand;
 import org.tensorflow.op.Ops;
-import org.tensorflow.types.family.TFloating;
+import org.tensorflow.types.family.TNumber;
 
 /**
  * Hyperbolic tangent activation function.
@@ -30,10 +30,8 @@
  *     Operand&lt;TFloat32&gt; result = tanh.call(input);
  *     // result = [-0.9950547f, -0.7615942f,  0.f,  0.7615942f,  0.9950547f]
  * </pre>
- *
- * @param <T> the data type of the activation
  */
-public class Tanh<T extends TFloating> extends Activation<T> {
+public class Tanh extends Activation {
 
   /**
    * Creates a Hyperbolic tangent activation.
@@ -46,7 +44,7 @@ public Tanh(Ops tf) {
 
   /** {@inheritDoc} */
   @Override
-  public Operand<T> call(Operand<T> input) {
+  public <T extends TNumber> Operand<T> call(Operand<T> input) {
     return tf.math.tanh(input);
   }
 }
diff --git a/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/BaseInitializer.java b/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/BaseInitializer.java
index 9c1fa9ac287..7efd02a6db1 100644
--- a/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/BaseInitializer.java
+++ b/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/BaseInitializer.java
@@ -15,10 +15,9 @@
 package org.tensorflow.framework.initializers;
 
 import org.tensorflow.op.Ops;
-import org.tensorflow.types.family.TType;
 
 /** Abstract base class for all Initializers */
-public abstract class BaseInitializer<T extends TType> implements Initializer<T> {
+public abstract class BaseInitializer implements Initializer {
 
   protected final Ops tf;
 
diff --git a/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Constant.java b/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Constant.java
index 4a2df86d74b..5aae5b90e5b 100644
--- a/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Constant.java
+++ b/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Constant.java
@@ -32,10 +32,8 @@
  *      Operand&lt;TFloat32&gt; values =
  *              initializer.call(tf.constant(Shape.of(2,2)), TFloat32.class);
  * </pre>
- *
- * @param <T> The Type for the call operation
  */
-public class Constant<T extends TType> extends BaseInitializer<T> {
+public class Constant extends BaseInitializer {
 
   private final double doubleValue;
   private final long longValue;
@@ -86,9 +84,10 @@ public Constant(Ops tf, boolean value) {
 
   /** {@inheritDoc} */
   @Override
-  public Operand<T> call(Operand<TInt64> dims, Class<T> type) {
+  public <T extends TType> Operand<T> call(Operand<TInt64> dims, Class<T> type) {
     if (!TNumber.class.isAssignableFrom(type) && type != TBool.class) {
-      throw new IllegalArgumentException("Tensor type must be numeric or boolean: " + type.getSimpleName());
+      throw new IllegalArgumentException(
+          "Tensor type must be numeric or boolean: " + type.getSimpleName());
     }
     switch (valueType) {
       case LONG:
diff --git a/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Glorot.java b/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Glorot.java
index 894bd073758..5a3c291785f 100644
--- a/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Glorot.java
+++ b/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Glorot.java
@@ -16,7 +16,6 @@
 package org.tensorflow.framework.initializers;
 
 import org.tensorflow.op.Ops;
-import org.tensorflow.types.family.TFloating;
 
 /**
  * The Glorot initializer, also called Xavier initializer.
@@ -58,16 +57,17 @@
  * </pre>
  *
  * <p><b>NOTE:</b>
+ *
  * <p>For a GlorotNormal equivalent initializer, use {@link
  * VarianceScaling.Distribution#TRUNCATED_NORMAL} for the distribution parameter.
+ *
  * <p>For a GlorotUniform equivalent initializer, use {@link VarianceScaling.Distribution#UNIFORM}
  * for the distribution parameter.
  *
- * @param <T> The TType for the call operation
  * @see VarianceScaling.Distribution
  * @see <a href="http://proceedings.mlr.press/v9/glorot10a.html">Glorot et al., 2010</a>
  */
-public class Glorot<T extends TFloating> extends VarianceScaling<T> {
+public class Glorot extends VarianceScaling {
 
   public static final double SCALE = 1.0;
 
@@ -77,7 +77,7 @@ public class Glorot<T extends TFloating> extends VarianceScaling<T> {
    * @param tf the TensorFlow Ops
    * @param distribution The distribution type for the Glorot initializer.
    * @param seed the seed for random number generation. An initializer created with a given seed
-   *     will always produce the same random tensor for a given shape and dtype.
+   *     will always produce the same random tensor for a given shape and data type.
    * @see VarianceScaling.Distribution
    */
   public Glorot(Ops tf, Distribution distribution, long seed) {
diff --git a/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/He.java b/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/He.java
index 3a91b72b0d0..ac64e449265 100644
--- a/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/He.java
+++ b/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/He.java
@@ -15,7 +15,6 @@
 package org.tensorflow.framework.initializers;
 
 import org.tensorflow.op.Ops;
-import org.tensorflow.types.family.TFloating;
 
 /**
  * He initializer.
@@ -53,17 +52,18 @@
  * </pre>
  *
  * <p><b>NOTE:</b>
+ *
  * <p>For an HeNormal equivalent initializer, use {@link
  * VarianceScaling.Distribution#TRUNCATED_NORMAL} for the distribution parameter.
- * <p>For an HeUniform equivalent initializer, use {@link VarianceScaling.Distribution#UNIFORM}
- * for the distribution parameter.
  *
- * @param <T> The TType for the call operation
+ * <p>For an HeUniform equivalent initializer, use {@link VarianceScaling.Distribution#UNIFORM} for
+ * the distribution parameter.
+ *
  * @see <a
  *     href="https://www.cv-foundation.org/openaccess/content_iccv_2015/html/He_Delving_Deep_into_ICCV_2015_paper.html">He
  *     et al., 2015</a>
  */
-public class He<T extends TFloating> extends VarianceScaling<T> {
+public class He extends VarianceScaling {
 
   public static final double SCALE = 2.0;
 
@@ -73,7 +73,7 @@ public class He<T extends TFloating> extends VarianceScaling<T> {
    * @param tf the TensorFlow Ops
    * @param distribution The distribution type for the He initializer.
    * @param seed the seed for random number generation. An initializer created with a given seed
-   *     will always produce the same random tensor for a given shape and dtype.
+   *     will always produce the same random tensor for a given shape and data type.
    * @see VarianceScaling.Distribution
    */
   public He(Ops tf, Distribution distribution, long seed) {
diff --git a/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Identity.java b/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Identity.java
index f672c9f1e85..e1f50e145a2 100644
--- a/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Identity.java
+++ b/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Identity.java
@@ -19,7 +19,7 @@
 import org.tensorflow.ndarray.Shape;
 import org.tensorflow.op.Ops;
 import org.tensorflow.types.TInt64;
-import org.tensorflow.types.family.TFloating;
+import org.tensorflow.types.family.TType;
 
 /**
  * Initializer that generates the identity matrix.
@@ -34,10 +34,8 @@
  *     Operand&lt;TFloat32&gt; values =
  *             initializer.call(tf.constant(Shape.of(2,2)), TFloat32.class);
  * </pre>
- *
- * @param <T> The TType for the call operation
  */
-public class Identity<T extends TFloating> extends BaseInitializer<T> {
+public class Identity extends BaseInitializer {
   public static final double GAIN_DEFAULT = 1.0;
 
   private final double gain;
@@ -65,7 +63,7 @@ public Identity(Ops tf, double gain) {
 
   /** {@inheritDoc} */
   @Override
-  public Operand<T> call(Operand<TInt64> dims, Class<T> type) {
+  public <T extends TType> Operand<T> call(Operand<TInt64> dims, Class<T> type) {
     Shape shape = ShapeUtils.toShape(tf.scope(), dims);
     if (shape.numDimensions() != 2) {
       throw new IllegalArgumentException("2D matrix required, got " + shape.numDimensions());
diff --git a/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Initializer.java b/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Initializer.java
index 4beb218783b..032f9c92792 100644
--- a/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Initializer.java
+++ b/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Initializer.java
@@ -18,12 +18,8 @@
 import org.tensorflow.types.TInt64;
 import org.tensorflow.types.family.TType;
 
-/**
- * An interface for Initializers
- *
- * @param <T> The data Type for initializer operation
- */
-public interface Initializer<T extends TType> {
+/** An interface for Initializers */
+public interface Initializer {
 
   /**
    * Generates the operation used to perform the initialization.
@@ -31,6 +27,7 @@ public interface Initializer<T extends TType> {
    * @param dims the shape dimensions
    * @param type the type of tensor
    * @return An operand for the initialization.
+   * @param <T> The data Type for initializer operation
    */
-  Operand<T> call(Operand<TInt64> dims, Class<T> type);
+  <T extends TType> Operand<T> call(Operand<TInt64> dims, Class<T> type);
 }
diff --git a/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/LeCun.java b/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/LeCun.java
index 38e68ef688b..b82f40918c0 100644
--- a/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/LeCun.java
+++ b/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/LeCun.java
@@ -15,7 +15,6 @@
 package org.tensorflow.framework.initializers;
 
 import org.tensorflow.op.Ops;
-import org.tensorflow.types.family.TFloating;
 
 /**
  * LeCun normal initializer.
@@ -27,7 +26,7 @@
  * stddev = sqrt(1 / fanIn)</code> where <code>fanIn</code> is the number of input units in the
  * weight tensor.
  *
- * <p>If the distribution is UNIFORM, itraws samples from a uniform distribution within <code>
+ * <p>If the distribution is UNIFORM, it draws samples from a uniform distribution within <code>
  * [-limit, limit]</code>, where <code>limit = Math.sqrt(3 / fanIn)</code> (<code>fanIn</code> is
  * the number of input units in the weight tensor)
  *
@@ -59,14 +58,14 @@
  *
  * <p><b>NOTE:</b> *
  *
- * <p>For a LeCunNormal equivalent initializer, use {@link VarianceScaling.Distribution#TRUNCATED_NORMAL} for the distribution parameter. *
+ * <p>For a LeCunNormal equivalent initializer, use {@link
+ * VarianceScaling.Distribution#TRUNCATED_NORMAL} for the distribution parameter. *
  *
  * <p>For a LeCunUniform equivalent initializer, use {@link VarianceScaling.Distribution#UNIFORM} *
  * for the distribution parameter. *
  *
  * <p>
  *
- * @param <T> The TType for the call operation
  * @see <a
  *     href="https://papers.nips.cc/paper/6698-self-normalizing-neural-networks">Self-Normalizing
  *     Neural Networks, Klambauer et al., 2017</a>
@@ -74,7 +73,7 @@
  *     al., 1998</a>
  * @see VarianceScaling.Distribution
  */
-public class LeCun<T extends TFloating> extends VarianceScaling<T> {
+public class LeCun extends VarianceScaling {
 
   /**
    * Creates a LeCunNormal Initializer
@@ -82,7 +81,7 @@ public class LeCun<T extends TFloating> extends VarianceScaling<T> {
    * @param tf the TensorFlow Ops
    * @param distribution The distribution type for the Glorot initializer.
    * @param seed the seed for random number generation. An initializer created with a given seed
-   *     will always produce the same random tensor for a given shape and dtype.
+   *     will always produce the same random tensor for a given shape and data type.
    */
   public LeCun(Ops tf, Distribution distribution, long seed) {
     super(tf, 1.0, Mode.FAN_IN, distribution, seed);
diff --git a/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Ones.java b/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Ones.java
index b8eb0c418e9..9094c2add97 100644
--- a/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Ones.java
+++ b/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Ones.java
@@ -32,10 +32,8 @@
  *      Operand&lt;TFloat32&gt; values =
  *              initializer.call(tf.constant(Shape.of(2,2)), TFloat32.class);
  * </pre>
- *
- * @param <T> The TType for the call operation
  */
-public class Ones<T extends TType> extends BaseInitializer<T> {
+public class Ones extends BaseInitializer {
 
   /**
    * Creates an Initializer that sets all values to one.
@@ -57,9 +55,10 @@ public Ones(Ops tf) {
 
   /** {@inheritDoc} */
   @Override
-  public Operand<T> call(Operand<TInt64> dims, Class<T> type) {
+  public <T extends TType> Operand<T> call(Operand<TInt64> dims, Class<T> type) {
     if (!TNumber.class.isAssignableFrom(type) && type != TBool.class) {
-      throw new IllegalArgumentException("Tensor type must be numeric or boolean: " + type.getSimpleName());
+      throw new IllegalArgumentException(
+          "Tensor type must be numeric or boolean: " + type.getSimpleName());
     }
     return tf.fill(dims, tf.dtypes.cast(tf.constant(1.0), type));
   }
diff --git a/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Orthogonal.java b/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Orthogonal.java
index a5b466e118e..b6dffd2d768 100644
--- a/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Orthogonal.java
+++ b/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Orthogonal.java
@@ -21,7 +21,10 @@
 import org.tensorflow.op.Ops;
 import org.tensorflow.op.linalg.Qr;
 import org.tensorflow.types.TInt64;
-import org.tensorflow.types.family.TFloating;
+import org.tensorflow.types.family.TNumber;
+import org.tensorflow.types.family.TType;
+
+import static org.tensorflow.framework.utils.CastHelper.cast;
 
 /**
  * Initializer that generates an orthogonal matrix.
@@ -44,10 +47,8 @@
  *      Operand&lt;TFloat32&gt; values =
  *              initializer.call(tf.constant(Shape.of(2,2)), TFloat32.class);
  * </pre>
- *
- * @param <T> The TType for the call operation
  */
-public class Orthogonal<T extends TFloating> extends BaseInitializer<T> {
+public class Orthogonal extends BaseInitializer {
 
   public static final double GAIN_DEFAULT = 1.0;
 
@@ -59,7 +60,7 @@ public class Orthogonal<T extends TFloating> extends BaseInitializer<T> {
    *
    * @param tf the TensorFlow Ops
    * @param seed the seed for random number generation. An initializer created with a given seed
-   *     will always produce the same random tensor for a given shape and dtype.
+   *     will always produce the same random tensor for a given shape and data type.
    */
   public Orthogonal(Ops tf, long seed) {
     this(tf, GAIN_DEFAULT, seed);
@@ -71,7 +72,7 @@ public Orthogonal(Ops tf, long seed) {
    * @param tf the TensorFlow Ops
    * @param gain the gain to be applied to the Matrix.
    * @param seed the seed for random number generation. An initializer created with a given seed
-   *     will always produce the same random tensor for a given shape and dtype.
+   *     will always produce the same random tensor for a given shape and data type.
    */
   public Orthogonal(Ops tf, double gain, long seed) {
     super(tf);
@@ -81,7 +82,11 @@ public Orthogonal(Ops tf, double gain, long seed) {
 
   /** {@inheritDoc} */
   @Override
-  public Operand<T> call(Operand<TInt64> dims, Class<T> type) {
+  public <T extends TType> Operand<T> call(Operand<TInt64> dims, Class<T> type) {
+    if (!TNumber.class.isAssignableFrom(type)) {
+      throw new IllegalArgumentException("Tensor type must be numeric: " + type.getSimpleName());
+    }
+    Class<TNumber> nType = (Class<TNumber>) type;
     Shape dimsShape = ShapeUtils.toShape(tf.scope(), dims);
     if (dimsShape.numDimensions() < 2) {
       throw new IllegalArgumentException(
@@ -94,17 +99,18 @@ public Operand<T> call(Operand<TInt64> dims, Class<T> type) {
     long numCols = dimsShape.size(i);
     Shape flatShape = Shape.of(Math.max(numRows, numCols), Math.min(numRows, numCols));
     long[] seeds = {seed, 0};
-    Operand<T> op =
-        tf.random.statelessRandomNormal(tf.constant(flatShape), tf.constant(seeds), type);
+    Operand<TNumber> op =
+        tf.random.statelessRandomNormal(tf.constant(flatShape), tf.constant(seeds), nType);
+
     Qr.Options qrOptions = Qr.fullMatrices(false);
-    Qr<T> qrOp = tf.linalg.qr(op, qrOptions);
-    Output<T> qo = qrOp.q();
-    Output<T> ro = qrOp.r();
-    Operand<T> diagOp =
-        tf.linalg.matrixDiagPart(ro, tf.constant(0), tf.dtypes.cast(tf.constant(0), type));
-    Operand<T> qop = tf.math.mul(qo, tf.math.sign(diagOp));
+    Qr<TNumber> qrOp = tf.linalg.qr(op, qrOptions);
+    Output<TNumber> qo = qrOp.q();
+    Output<TNumber> ro = qrOp.r();
+    Operand<TNumber> diagOp =
+        tf.linalg.matrixDiagPart(ro, tf.constant(0), tf.dtypes.cast(tf.constant(0), op.type()));
+    Operand<TNumber> qop = tf.math.mul(qo, tf.math.sign(diagOp));
     if (numRows < numCols) qop = tf.linalg.transpose(qop, null);
 
-    return tf.math.mul(qop, tf.dtypes.cast(tf.constant(this.gain), type));
+    return cast(tf, tf.math.mul(qop, tf.dtypes.cast(tf.constant(this.gain), op.type())), type);
   }
 }
diff --git a/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/RandomNormal.java b/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/RandomNormal.java
index 38ab194a56b..8f8edcc56ca 100644
--- a/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/RandomNormal.java
+++ b/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/RandomNormal.java
@@ -17,7 +17,10 @@
 import org.tensorflow.Operand;
 import org.tensorflow.op.Ops;
 import org.tensorflow.types.TInt64;
-import org.tensorflow.types.family.TFloating;
+import org.tensorflow.types.family.TNumber;
+import org.tensorflow.types.family.TType;
+
+import static org.tensorflow.framework.utils.CastHelper.cast;
 
 /**
  * Initializer that generates tensors with a normal distribution.
@@ -31,10 +34,8 @@
  *     Operand&lt;TFloat32&gt; values =
  *              initializer.call(tf.constant(Shape.of(2,2)), TFloat32.class);
  * </pre>
- *
- * @param <T> The TType for the call operation
  */
-public class RandomNormal<T extends TFloating> extends BaseInitializer<T> {
+public class RandomNormal extends BaseInitializer {
 
   public static final double MEAN_DEFAULT = 0.0;
   public static final double STDDEV_DEFAULT = 1.0;
@@ -49,7 +50,7 @@ public class RandomNormal<T extends TFloating> extends BaseInitializer<T> {
    *
    * @param tf the TensorFlow Ops
    * @param seed the seed for random number generation. An initializer created with a given seed
-   *     will always produce the same random tensor for a given shape and dtype.
+   *     will always produce the same random tensor for a given shape and data type.
    */
   public RandomNormal(Ops tf, long seed) {
     this(tf, MEAN_DEFAULT, STDDEV_DEFAULT, seed);
@@ -61,7 +62,7 @@ public RandomNormal(Ops tf, long seed) {
    * @param tf the TensorFlow Ops
    * @param mean Mean of the random values to generate.
    * @param seed the seed for random number generation. An initializer created with a given seed
-   *     will always produce the same random tensor for a given shape and dtype.
+   *     will always produce the same random tensor for a given shape and data type.
    */
   public RandomNormal(Ops tf, double mean, long seed) {
     this(tf, mean, STDDEV_DEFAULT, seed);
@@ -74,7 +75,7 @@ public RandomNormal(Ops tf, double mean, long seed) {
    * @param mean Mean of the random values to generate.
    * @param stddev Standard deviation of the random values to generate.
    * @param seed the seed for random number generation. An initializer created with a given seed
-   *     will always produce the same random tensor for a given shape and dtype.
+   *     will always produce the same random tensor for a given shape and data type.
    */
   public RandomNormal(Ops tf, double mean, double stddev, long seed) {
     super(tf);
@@ -85,10 +86,16 @@ public RandomNormal(Ops tf, double mean, double stddev, long seed) {
 
   /** {@inheritDoc} */
   @Override
-  public Operand<T> call(Operand<TInt64> dims, Class<T> type) {
+  public <T extends TType> Operand<T> call(Operand<TInt64> dims, Class<T> type) {
+    if (!TNumber.class.isAssignableFrom(type)) {
+      throw new IllegalArgumentException("Tensor type must be numeric: " + type.getSimpleName());
+    }
     long[] seeds = {seed, 0};
-    Operand<T> distOp = tf.random.statelessRandomNormal(dims, tf.constant(seeds), type);
-    Operand<T> op = tf.math.mul(distOp, tf.dtypes.cast(tf.constant(this.stddev), type));
-    return tf.math.add(op, tf.dtypes.cast(tf.constant(mean), type));
+    @SuppressWarnings("unchecked")
+    Class<TNumber> nType = (Class<TNumber>) type;
+    Operand<TNumber> distOp = tf.random.statelessRandomNormal(dims, tf.constant(seeds), nType);
+    Operand<TNumber> op =
+        tf.math.mul(distOp, tf.dtypes.cast(tf.constant(this.stddev), distOp.type()));
+    return cast(tf, tf.math.add(op, tf.dtypes.cast(tf.constant(mean), distOp.type())), type);
   }
 }
diff --git a/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/RandomUniform.java b/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/RandomUniform.java
index 787af15f709..3c7d394d538 100644
--- a/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/RandomUniform.java
+++ b/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/RandomUniform.java
@@ -20,6 +20,9 @@
 import org.tensorflow.types.TInt64;
 import org.tensorflow.types.family.TIntegral;
 import org.tensorflow.types.family.TNumber;
+import org.tensorflow.types.family.TType;
+
+import static org.tensorflow.framework.utils.CastHelper.cast;
 
 /**
  * Initializer that generates tensors with a uniform distribution.
@@ -33,10 +36,8 @@
  *     Operand&lt;TFloat32&gt; values =
  *              initializer.call(tf.constant(Shape.of(2,2)), TFloat32.class);
  * </pre>
- *
- * @param <T> The TType for the call operation
  */
-public class RandomUniform<T extends TNumber> extends BaseInitializer<T> {
+public class RandomUniform extends BaseInitializer {
 
   public static final double MINVAL_DEFAULT = -0.05;
   public static final double MAXVAL_DEFAULT = 0.05;
@@ -46,12 +47,12 @@ public class RandomUniform<T extends TNumber> extends BaseInitializer<T> {
   private final long seed;
 
   /**
-   * Creates a RandomUniform initializer using {@link #MINVAL_DEFAULT} for the minval and
-   * {@link #MAXVAL_DEFAULT} for the maxval
+   * Creates a RandomUniform initializer using {@link #MINVAL_DEFAULT} for the minval and {@link
+   * #MAXVAL_DEFAULT} for the maxval
    *
    * @param tf the TensorFlow Ops
    * @param seed the seed for random number generation. An initializer created with a given seed
-   *     will always produce the same random tensor for a given shape and dtype.
+   *     will always produce the same random tensor for a given shape and data type.
    */
   public RandomUniform(Ops tf, long seed) {
     this(tf, MINVAL_DEFAULT, MAXVAL_DEFAULT, seed);
@@ -64,7 +65,7 @@ public RandomUniform(Ops tf, long seed) {
    * @param minval Lower bound of the range of random values to generate (inclusive).
    * @param maxval Upper bound of the range of random values to generate (exclusive).
    * @param seed the seed for random number generation. An initializer created with a given seed
-   *     will always produce the same random tensor for a given shape and dtype.
+   *     will always produce the same random tensor for a given shape and data type.
    */
   public RandomUniform(Ops tf, double minval, double maxval, long seed) {
     super(tf);
@@ -75,28 +76,34 @@ public RandomUniform(Ops tf, double minval, double maxval, long seed) {
 
   /** {@inheritDoc} */
   @Override
-  public Operand<T> call(Operand<TInt64> dims, Class<T> type) {
-    Operand<T> distOp;
+  public <T extends TType> Operand<T> call(Operand<TInt64> dims, Class<T> type) {
+    if (!TNumber.class.isAssignableFrom(type)) {
+      throw new IllegalArgumentException("Tensor type must be numeric: " + type.getSimpleName());
+    }
+    @SuppressWarnings("unchecked")
+    Class<TNumber> nType = (Class<TNumber>) type;
+    Operand<TNumber> distOp;
     if (TIntegral.class.isAssignableFrom(type)) {
       RandomUniformInt.Options options = RandomUniformInt.seed(this.seed);
       distOp =
           tf.random.randomUniformInt(
               dims,
-              tf.dtypes.cast(tf.constant(this.minval), type),
-              tf.dtypes.cast(tf.constant(this.maxval), type),
+              cast(tf, tf.constant(this.minval), nType),
+              cast(tf, tf.constant(this.maxval), nType),
               options);
     } else {
       long[] seeds = {seed, 0};
-      distOp = tf.random.statelessRandomUniform(dims, tf.constant(seeds), type);
+      distOp = tf.random.statelessRandomUniform(dims, tf.constant(seeds), nType);
       if (this.minval == 0) {
         if (this.maxval != 1.0) {
-          distOp = tf.math.mul(distOp, tf.dtypes.cast(tf.constant(this.maxval), type));
+          distOp = tf.math.mul(distOp, cast(tf, tf.constant(this.maxval), distOp.type()));
         }
       } else {
-        distOp = tf.math.mul(distOp, tf.dtypes.cast(tf.constant(this.maxval - this.minval), type));
-        distOp = tf.math.add(distOp, tf.dtypes.cast(tf.constant(this.minval), type));
+        distOp =
+            tf.math.mul(distOp, cast(tf, tf.constant(this.maxval - this.minval), distOp.type()));
+        distOp = tf.math.add(distOp, cast(tf, tf.constant(this.minval), distOp.type()));
       }
     }
-    return distOp;
+    return cast(tf, distOp, type);
   }
 }
diff --git a/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/TruncatedNormal.java b/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/TruncatedNormal.java
index d3cfec26338..7969e6988fa 100644
--- a/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/TruncatedNormal.java
+++ b/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/TruncatedNormal.java
@@ -17,7 +17,10 @@
 import org.tensorflow.Operand;
 import org.tensorflow.op.Ops;
 import org.tensorflow.types.TInt64;
-import org.tensorflow.types.family.TFloating;
+import org.tensorflow.types.family.TNumber;
+import org.tensorflow.types.family.TType;
+
+import static org.tensorflow.framework.utils.CastHelper.cast;
 
 /**
  * Initializer that generates a truncated normal distribution.
@@ -31,10 +34,8 @@
  *     Operand&lt;TFloat32&gt; values =
  *              initializer.call(tf.constant(Shape.of(2,2)), TFloat32.class);
  * </pre>
- *
- * @param <T> The TType for the call operation
  */
-public class TruncatedNormal<T extends TFloating> extends BaseInitializer<T> {
+public class TruncatedNormal extends BaseInitializer {
 
   public static final double MEAN_DEFAULT = 0.0;
   public static final double STDDEV_DEFAULT = 0.05;
@@ -49,7 +50,7 @@ public class TruncatedNormal<T extends TFloating> extends BaseInitializer<T> {
    *
    * @param tf the TensorFlow Ops
    * @param seed the seed for random number generation. An initializer created with a given seed
-   *     will always produce the same random tensor for a given shape and dtype.
+   *     will always produce the same random tensor for a given shape and data type.
    */
   public TruncatedNormal(Ops tf, long seed) {
     this(tf, MEAN_DEFAULT, STDDEV_DEFAULT, seed);
@@ -62,7 +63,7 @@ public TruncatedNormal(Ops tf, long seed) {
    * @param mean Mean of the random values to generate.
    * @param stddev Standard deviation of the random values to generate.
    * @param seed the seed for random number generation. An initializer created with a given seed
-   *     will always produce the same random tensor for a given shape and dtype.
+   *     will always produce the same random tensor for a given shape and data type.
    */
   public TruncatedNormal(Ops tf, double mean, double stddev, long seed) {
     super(tf);
@@ -73,11 +74,19 @@ public TruncatedNormal(Ops tf, double mean, double stddev, long seed) {
 
   /** {@inheritDoc} */
   @Override
-  public Operand<T> call(Operand<TInt64> dims, Class<T> type) {
-    long[] seeds = {seed,0};
-    Operand<T> distOp = tf.random.statelessTruncatedNormal(dims, tf.constant(seeds), type);
-    return tf.math.add(
-        tf.math.mul(distOp, tf.dtypes.cast(tf.constant(stddev), type)),
-        tf.dtypes.cast(tf.constant(mean), type));
+  public <T extends TType> Operand<T> call(Operand<TInt64> dims, Class<T> type) {
+    if (!TNumber.class.isAssignableFrom(type)) {
+      throw new IllegalArgumentException("Tensor type must be numeric: " + type.getSimpleName());
+    }
+    @SuppressWarnings("unchecked")
+    Class<TNumber> nType = (Class<TNumber>) type;
+    long[] seeds = {seed, 0};
+    Operand<TNumber> distOp = tf.random.statelessTruncatedNormal(dims, tf.constant(seeds), nType);
+    return cast(
+        tf,
+        tf.math.add(
+            tf.math.mul(distOp, cast(tf, tf.constant(stddev), distOp.type())),
+            cast(tf, tf.constant(mean), distOp.type())),
+        type);
   }
 }
diff --git a/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/VarianceScaling.java b/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/VarianceScaling.java
index 5d951450505..583d5680c57 100644
--- a/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/VarianceScaling.java
+++ b/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/VarianceScaling.java
@@ -19,13 +19,16 @@
 import org.tensorflow.ndarray.Shape;
 import org.tensorflow.op.Ops;
 import org.tensorflow.types.TInt64;
-import org.tensorflow.types.family.TFloating;
+import org.tensorflow.types.family.TNumber;
+import org.tensorflow.types.family.TType;
+
+import static org.tensorflow.framework.utils.CastHelper.cast;
 
 /**
  * Initializer capable of adapting its scale to the shape of weights tensors.
  *
- * <p>With <code>distribution=TRUNCATED_NORMAL or NORMAL</code>, samples are drawn from
- * a truncated/untruncated normal distribution with a mean of zero and a standard deviation (after
+ * <p>With <code>distribution=TRUNCATED_NORMAL or NORMAL</code>, samples are drawn from a
+ * truncated/untruncated normal distribution with a mean of zero and a standard deviation (after
  * truncation, if used) <code>stddev = Math.sqrt(scale / n)</code>, where <code>n</code> is:
  *
  * <ul>
@@ -49,11 +52,10 @@
  *          initializer.call(tf.constant(Shape.of(2,2)), TFloat32.class);
  * </pre>
  *
- * @param <T> The TType for the call operation
  * @see VarianceScaling.Mode
  * @see VarianceScaling.Distribution
  */
-public class VarianceScaling<T extends TFloating> extends BaseInitializer<T> {
+public class VarianceScaling extends BaseInitializer {
 
   public static final double SCALE_DEFAULT = 1.0;
   public static final Mode MODE_DEFAULT = Mode.FAN_IN;
@@ -64,7 +66,6 @@ public class VarianceScaling<T extends TFloating> extends BaseInitializer<T> {
   private final Distribution distribution;
   private final long seed;
 
-
   /**
    * Creates a VarianceScaling Initializer
    *
@@ -97,7 +98,12 @@ public VarianceScaling(Ops tf, double scale, Mode mode, Distribution distributio
 
   /** {@inheritDoc} */
   @Override
-  public Operand<T> call(Operand<TInt64> dims, Class<T> type) {
+  public <T extends TType> Operand<T> call(Operand<TInt64> dims, Class<T> type) {
+    if (!TNumber.class.isAssignableFrom(type)) {
+      throw new IllegalArgumentException("Tensor type must be numeric: " + type.getSimpleName());
+    }
+    @SuppressWarnings("unchecked")
+    Class<TNumber> nType = (Class<TNumber>) type;
     Shape shape = ShapeUtils.toShape(this.tf.scope(), dims);
     double lscale = this.scale;
     double[] fans /* fanIn, fanOut */ = computeFans(shape);
@@ -112,28 +118,28 @@ public Operand<T> call(Operand<TInt64> dims, Class<T> type) {
         lscale /= Math.max(1., (fans[0] + fans[1]) / 2.);
         break;
     }
-    Operand<T> distOp;
-    Operand<T> mulOp = null;
+    Operand<TNumber> distOp;
+    Operand<TNumber> mulOp = null;
     double stddev;
     long[] seeds = {seed, 0};
     switch (distribution) {
       case TRUNCATED_NORMAL:
-        distOp = tf.random.statelessTruncatedNormal(dims, tf.constant(seeds), type);
+        distOp = tf.random.statelessTruncatedNormal(dims, tf.constant(seeds), nType);
         stddev = Math.sqrt(lscale) / .87962566103423978;
-        mulOp = tf.math.mul(distOp, tf.dtypes.cast(tf.constant(stddev), type));
+        mulOp = tf.math.mul(distOp, cast(tf, tf.constant(stddev), nType));
         break;
       case NORMAL:
-        distOp = tf.random.statelessRandomNormal(dims, tf.constant(seeds), type);
+        distOp = tf.random.statelessRandomNormal(dims, tf.constant(seeds), nType);
         stddev = Math.sqrt(lscale);
-        mulOp = tf.math.mul(distOp, tf.dtypes.cast(tf.constant(stddev), type));
+        mulOp = tf.math.mul(distOp, cast(tf, tf.constant(stddev), nType));
         break;
       case UNIFORM:
-        distOp = tf.random.statelessRandomUniform(dims, tf.constant(seeds), type);
+        distOp = tf.random.statelessRandomUniform(dims, tf.constant(seeds), nType);
         stddev = Math.sqrt(3.0 * lscale);
-        mulOp = tf.math.mul(distOp, tf.dtypes.cast(tf.constant(stddev), type));
+        mulOp = tf.math.mul(distOp, cast(tf, tf.constant(stddev), nType));
         break;
     }
-    return mulOp;
+    return cast(tf, mulOp, type);
   }
 
   /**
diff --git a/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Zeros.java b/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Zeros.java
index 4298493ac44..1f8c08c7910 100644
--- a/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Zeros.java
+++ b/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Zeros.java
@@ -30,10 +30,8 @@
  *      Operand&lt;TFloat32&gt; values =
  *              initializer.call(tf.constant(Shape.of(2,2)), TFloat32.class);
  * </pre>
- *
- * @param <T> The TType for the call operation
  */
-public class Zeros<T extends TType> extends BaseInitializer<T> {
+public class Zeros extends BaseInitializer {
 
   /**
    * Creates an Initializer that sets all values to one.
@@ -45,7 +43,7 @@ public Zeros(Ops tf) {
   }
 
   @Override
-  public Operand<T> call(Operand<TInt64> dims, Class<T> dtype) {
-    return tf.zeros(dims, dtype);
+  public <T extends TType> Operand<T> call(Operand<TInt64> dims, Class<T> type) {
+    return tf.zeros(dims, type);
   }
 }
diff --git a/tensorflow-framework/src/test/java/org/tensorflow/framework/activations/ELUTest.java b/tensorflow-framework/src/test/java/org/tensorflow/framework/activations/ELUTest.java
index 914b94dfada..22604a02c08 100644
--- a/tensorflow-framework/src/test/java/org/tensorflow/framework/activations/ELUTest.java
+++ b/tensorflow-framework/src/test/java/org/tensorflow/framework/activations/ELUTest.java
@@ -21,8 +21,6 @@
 import org.tensorflow.types.TFloat32;
 import org.tensorflow.types.TFloat64;
 
-import static org.junit.jupiter.api.Assertions.assertThrows;
-
 /** @author Jim Clarke */
 public class ELUTest {
 
@@ -42,8 +40,6 @@ public void setUp() {}
   @AfterEach
   public void tearDown() {}
 
-
-
   /** Test of ELU call method */
   @Test
   public void testCallFloat() {
@@ -52,7 +48,7 @@ public void testCallFloat() {
     for (TestSession.Mode tfMode : tfModes)
       try (TestSession session = TestSession.createTestSession(tfMode)) {
         Ops tf = session.getTF();
-        ELU<TFloat32> instance = new ELU<>(tf);
+        ELU instance = new ELU(tf);
         Operand<TFloat32> result = instance.call(tf.constant(input));
         session.evaluate(expected, result);
       }
@@ -66,7 +62,7 @@ public void testCallDouble() {
     for (TestSession.Mode tfMode : tfModes)
       try (TestSession session = TestSession.createTestSession(tfMode)) {
         Ops tf = session.getTF();
-        ELU<TFloat64> instance = new ELU<>(tf);
+        ELU instance = new ELU(tf);
         Operand<TFloat64> result = instance.call(tf.constant(input));
         session.evaluate(expected, result);
       }
@@ -80,7 +76,7 @@ public void testAlpha() {
     for (TestSession.Mode tfMode : tfModes)
       try (TestSession session = TestSession.createTestSession(tfMode)) {
         Ops tf = session.getTF();
-        ELU<TFloat64> instance = new ELU<>(tf, 2.0f);
+        ELU instance = new ELU(tf, 2.0f);
         Operand<TFloat64> result = instance.call(tf.constant(input));
         session.evaluate(expected, result);
       }
diff --git a/tensorflow-framework/src/test/java/org/tensorflow/framework/activations/ExponentialTest.java b/tensorflow-framework/src/test/java/org/tensorflow/framework/activations/ExponentialTest.java
index 1157c582168..e8758148c21 100644
--- a/tensorflow-framework/src/test/java/org/tensorflow/framework/activations/ExponentialTest.java
+++ b/tensorflow-framework/src/test/java/org/tensorflow/framework/activations/ExponentialTest.java
@@ -21,8 +21,6 @@
 import org.tensorflow.types.TFloat32;
 import org.tensorflow.types.TFloat64;
 
-import static org.junit.jupiter.api.Assertions.assertThrows;
-
 /** @author Jim Clarke */
 public class ExponentialTest {
   private final TestSession.Mode[] tfModes = {TestSession.Mode.EAGER, TestSession.Mode.GRAPH};
@@ -41,8 +39,6 @@ public void setUp() {}
   @AfterEach
   public void tearDown() {}
 
-
-
   /** Test of Exponential call method. */
   @Test
   public void testCallFloat() {
@@ -60,7 +56,7 @@ public void testCallFloat() {
     for (TestSession.Mode tfMode : tfModes)
       try (TestSession session = TestSession.createTestSession(tfMode)) {
         Ops tf = session.getTF();
-        Exponential<TFloat32> instance = new Exponential<>(tf);
+        Exponential instance = new Exponential(tf);
         Operand<TFloat32> result = instance.call(tf.constant(input));
         session.evaluate(expected, result);
       }
@@ -78,7 +74,7 @@ public void testCallDouble() {
     for (TestSession.Mode tfMode : tfModes)
       try (TestSession session = TestSession.createTestSession(tfMode)) {
         Ops tf = session.getTF();
-        Exponential<TFloat64> instance = new Exponential<>(tf);
+        Exponential instance = new Exponential(tf);
         Operand<TFloat64> result = instance.call(tf.constant(input));
         session.evaluate(expected, result);
       }
diff --git a/tensorflow-framework/src/test/java/org/tensorflow/framework/activations/HardSigmoidTest.java b/tensorflow-framework/src/test/java/org/tensorflow/framework/activations/HardSigmoidTest.java
index 35f57c47f66..59370c745a3 100644
--- a/tensorflow-framework/src/test/java/org/tensorflow/framework/activations/HardSigmoidTest.java
+++ b/tensorflow-framework/src/test/java/org/tensorflow/framework/activations/HardSigmoidTest.java
@@ -21,8 +21,6 @@
 import org.tensorflow.types.TFloat32;
 import org.tensorflow.types.TFloat64;
 
-import static org.junit.jupiter.api.Assertions.assertThrows;
-
 /** @author Jim Clarke */
 public class HardSigmoidTest {
   private final TestSession.Mode[] tfModes = {TestSession.Mode.EAGER, TestSession.Mode.GRAPH};
@@ -41,8 +39,6 @@ public void setUp() {}
   @AfterEach
   public void tearDown() {}
 
-
-
   /** Test of HardSigmoid call method. */
   @Test
   public void testCallFloat() {
@@ -51,7 +47,7 @@ public void testCallFloat() {
     for (TestSession.Mode tfMode : tfModes)
       try (TestSession session = TestSession.createTestSession(tfMode)) {
         Ops tf = session.getTF();
-        HardSigmoid<TFloat32> instance = new HardSigmoid<>(tf);
+        HardSigmoid instance = new HardSigmoid(tf);
         Operand<TFloat32> result = instance.call(tf.constant(input));
         session.evaluate(expected, result);
       }
@@ -65,7 +61,7 @@ public void testCallDouble() {
     for (TestSession.Mode tfMode : tfModes)
       try (TestSession session = TestSession.createTestSession(tfMode)) {
         Ops tf = session.getTF();
-        HardSigmoid<TFloat64> instance = new HardSigmoid<>(tf);
+        HardSigmoid instance = new HardSigmoid(tf);
         Operand<TFloat64> result = instance.call(tf.constant(input));
         session.evaluate(expected, result);
       }
diff --git a/tensorflow-framework/src/test/java/org/tensorflow/framework/activations/LinearTest.java b/tensorflow-framework/src/test/java/org/tensorflow/framework/activations/LinearTest.java
index 7974035c680..4e82420c922 100644
--- a/tensorflow-framework/src/test/java/org/tensorflow/framework/activations/LinearTest.java
+++ b/tensorflow-framework/src/test/java/org/tensorflow/framework/activations/LinearTest.java
@@ -48,7 +48,7 @@ public void testCallInt() {
     for (TestSession.Mode tfMode : tfModes)
       try (TestSession session = TestSession.createTestSession(tfMode)) {
         Ops tf = session.getTF();
-        Linear<TInt32> instance = new Linear<>(tf);
+        Linear instance = new Linear(tf);
         Operand<TInt32> result = instance.call(tf.constant(input));
         session.evaluate(expected, result);
       }
@@ -62,7 +62,7 @@ public void testCallFloat() {
     for (TestSession.Mode tfMode : tfModes)
       try (TestSession session = TestSession.createTestSession(tfMode)) {
         Ops tf = session.getTF();
-        Linear<TFloat32> instance = new Linear<>(tf);
+        Linear instance = new Linear(tf);
         Operand<TFloat32> result = instance.call(tf.constant(input));
         session.evaluate(expected, result);
       }
@@ -76,7 +76,7 @@ public void testCallDouble() {
     for (TestSession.Mode tfMode : tfModes)
       try (TestSession session = TestSession.createTestSession(tfMode)) {
         Ops tf = session.getTF();
-        Linear<TFloat64> instance = new Linear<>(tf);
+        Linear instance = new Linear(tf);
         Operand<TFloat64> result = instance.call(tf.constant(input));
         session.evaluate(expected, result);
       }
diff --git a/tensorflow-framework/src/test/java/org/tensorflow/framework/activations/ReLUTest.java b/tensorflow-framework/src/test/java/org/tensorflow/framework/activations/ReLUTest.java
index a0aa2c4b453..3eb32acff86 100644
--- a/tensorflow-framework/src/test/java/org/tensorflow/framework/activations/ReLUTest.java
+++ b/tensorflow-framework/src/test/java/org/tensorflow/framework/activations/ReLUTest.java
@@ -46,7 +46,7 @@ public void testCallFloat() {
     for (TestSession.Mode tfMode : tfModes)
       try (TestSession session = TestSession.createTestSession(tfMode)) {
         Ops tf = session.getTF();
-        ReLU<TFloat32> instance = new ReLU<>(tf);
+        ReLU instance = new ReLU(tf);
         Operand<TFloat32> result = instance.call(tf.constant(input));
         session.evaluate(tf.constant(expected), result);
       }
@@ -60,7 +60,7 @@ public void testCallInt() {
     for (TestSession.Mode tfMode : tfModes)
       try (TestSession session = TestSession.createTestSession(tfMode)) {
         Ops tf = session.getTF();
-        ReLU<TInt32> instance = new ReLU<>(tf);
+        ReLU instance = new ReLU(tf);
         Operand<TInt32> result = instance.call(tf.constant(input));
         session.evaluate(tf.constant(expected), result);
       }
@@ -74,7 +74,7 @@ public void testCallLong() {
     for (TestSession.Mode tfMode : tfModes)
       try (TestSession session = TestSession.createTestSession(tfMode)) {
         Ops tf = session.getTF();
-        ReLU<TInt64> instance = new ReLU<>(tf);
+        ReLU instance = new ReLU(tf);
         Operand<TInt64> result = instance.call(tf.constant(input));
         session.evaluate(tf.constant(expected), result);
       }
@@ -88,7 +88,7 @@ public void testCallFloat16() {
     for (TestSession.Mode tfMode : tfModes)
       try (TestSession session = TestSession.createTestSession(tfMode)) {
         Ops tf = session.getTF();
-        ReLU<TFloat16> instance = new ReLU<>(tf);
+        ReLU instance = new ReLU(tf);
         Operand<TFloat16> result =
             instance.call(tf.dtypes.cast(tf.constant(input), TFloat16.class));
         session.evaluate(tf.dtypes.cast(tf.constant(expected), TFloat16.class), result);
@@ -103,7 +103,7 @@ public void testCallDouble() {
     for (TestSession.Mode tfMode : tfModes)
       try (TestSession session = TestSession.createTestSession(tfMode)) {
         Ops tf = session.getTF();
-        ReLU<TFloat64> instance = new ReLU<>(tf);
+        ReLU instance = new ReLU(tf);
         Operand<TFloat64> result = instance.call(tf.constant(input));
         session.evaluate(tf.constant(expected), result);
       }
@@ -112,11 +112,11 @@ public void testCallDouble() {
   @Test
   public void testAlpha() {
     double[] input = {-10., -5., 0.0, 5., 10.};
-    double[] expected = {-5. , -2.5,  0.,  5., 10.};
+    double[] expected = {-5., -2.5, 0., 5., 10.};
     for (TestSession.Mode tfMode : tfModes)
       try (TestSession session = TestSession.createTestSession(tfMode)) {
         Ops tf = session.getTF();
-        ReLU<TFloat64> instance = new ReLU<>(tf, 0.5f, ReLU.MAX_VALUE_DEFAULT, ReLU.THRESHOLD_DEFAULT);
+        ReLU instance = new ReLU(tf, 0.5f, ReLU.MAX_VALUE_DEFAULT, ReLU.THRESHOLD_DEFAULT);
         Operand<TFloat64> result = instance.call(tf.constant(input));
         session.evaluate(tf.constant(expected), result);
       }
@@ -129,7 +129,7 @@ public void testMaxValue() {
     for (TestSession.Mode tfMode : tfModes)
       try (TestSession session = TestSession.createTestSession(tfMode)) {
         Ops tf = session.getTF();
-        ReLU<TFloat64> instance = new ReLU<>(tf, ReLU.ALPHA_DEFAULT, 5, ReLU.THRESHOLD_DEFAULT);
+        ReLU instance = new ReLU(tf, ReLU.ALPHA_DEFAULT, 5, ReLU.THRESHOLD_DEFAULT);
         Operand<TFloat64> result = instance.call(tf.constant(input));
         session.evaluate(tf.constant(expected), result);
       }
@@ -138,11 +138,11 @@ public void testMaxValue() {
   @Test
   public void testThreshold() {
     double[] input = {-10., -5., 0.0, 5., 10.};
-    double[] expected = {-0., -0.,  0.,  0., 10.};
+    double[] expected = {-0., -0., 0., 0., 10.};
     for (TestSession.Mode tfMode : tfModes)
       try (TestSession session = TestSession.createTestSession(tfMode)) {
         Ops tf = session.getTF();
-        ReLU<TFloat64> instance = new ReLU<>(tf, ReLU.ALPHA_DEFAULT, ReLU.MAX_VALUE_DEFAULT, 5.0f);
+        ReLU instance = new ReLU(tf, ReLU.ALPHA_DEFAULT, ReLU.MAX_VALUE_DEFAULT, 5.0f);
         Operand<TFloat64> result = instance.call(tf.constant(input));
         session.evaluate(tf.constant(expected), result);
       }
diff --git a/tensorflow-framework/src/test/java/org/tensorflow/framework/activations/SELUTest.java b/tensorflow-framework/src/test/java/org/tensorflow/framework/activations/SELUTest.java
index 8bad6f1f066..65d1a2f0135 100644
--- a/tensorflow-framework/src/test/java/org/tensorflow/framework/activations/SELUTest.java
+++ b/tensorflow-framework/src/test/java/org/tensorflow/framework/activations/SELUTest.java
@@ -21,8 +21,6 @@
 import org.tensorflow.types.TFloat32;
 import org.tensorflow.types.TFloat64;
 
-import static org.junit.jupiter.api.Assertions.assertThrows;
-
 /** @author Jim Clarke */
 public class SELUTest {
   private final TestSession.Mode[] tfModes = {TestSession.Mode.EAGER, TestSession.Mode.GRAPH};
@@ -41,8 +39,6 @@ public void setUp() {}
   @AfterEach
   public void tearDown() {}
 
-
-
   /** Test of SELU call method */
   @Test
   public void testCallFloat() {
@@ -53,7 +49,7 @@ public void testCallFloat() {
     for (TestSession.Mode tfMode : tfModes)
       try (TestSession session = TestSession.createTestSession(tfMode)) {
         Ops tf = session.getTF();
-        SELU<TFloat32> instance = new SELU<>(tf);
+        SELU instance = new SELU(tf);
         Operand<TFloat32> result = instance.call(tf.constant(input));
         session.evaluate(expected, result);
       }
@@ -71,7 +67,7 @@ public void testCallDouble() {
     for (TestSession.Mode tfMode : tfModes)
       try (TestSession session = TestSession.createTestSession(tfMode)) {
         Ops tf = session.getTF();
-        SELU<TFloat64> instance = new SELU<>(tf);
+        SELU instance = new SELU(tf);
         Operand<TFloat64> result = instance.call(tf.constant(input));
         session.evaluate(expected, result);
       }
diff --git a/tensorflow-framework/src/test/java/org/tensorflow/framework/activations/SigmoidTest.java b/tensorflow-framework/src/test/java/org/tensorflow/framework/activations/SigmoidTest.java
index 9dca622c3ec..6177d142794 100644
--- a/tensorflow-framework/src/test/java/org/tensorflow/framework/activations/SigmoidTest.java
+++ b/tensorflow-framework/src/test/java/org/tensorflow/framework/activations/SigmoidTest.java
@@ -21,8 +21,6 @@
 import org.tensorflow.types.TFloat32;
 import org.tensorflow.types.TFloat64;
 
-import static org.junit.jupiter.api.Assertions.assertThrows;
-
 /** @author Jim Clarke */
 public class SigmoidTest {
   private final TestSession.Mode[] tfModes = {TestSession.Mode.EAGER, TestSession.Mode.GRAPH};
@@ -41,7 +39,6 @@ public void setUp() {}
   @AfterEach
   public void tearDown() {}
 
-
   /** Test of Sigmoid call method */
   @Test
   public void testCallFloat() {
@@ -59,7 +56,7 @@ public void testCallFloat() {
     for (TestSession.Mode tfMode : tfModes)
       try (TestSession session = TestSession.createTestSession(tfMode)) {
         Ops tf = session.getTF();
-        Sigmoid<TFloat32> instance = new Sigmoid<>(tf);
+        Sigmoid instance = new Sigmoid(tf);
         Operand<TFloat32> result = instance.call(tf.constant(input));
         session.evaluate(expected, result);
       }
@@ -77,7 +74,7 @@ public void testCallDouble() {
     for (TestSession.Mode tfMode : tfModes)
       try (TestSession session = TestSession.createTestSession(tfMode)) {
         Ops tf = session.getTF();
-        Sigmoid<TFloat64> instance = new Sigmoid<>(tf);
+        Sigmoid instance = new Sigmoid(tf);
         Operand<TFloat64> result = instance.call(tf.constant(input));
         session.evaluate(expected, result);
       }
diff --git a/tensorflow-framework/src/test/java/org/tensorflow/framework/activations/SoftmaxTest.java b/tensorflow-framework/src/test/java/org/tensorflow/framework/activations/SoftmaxTest.java
index 05ec3a4f716..749b8f34602 100644
--- a/tensorflow-framework/src/test/java/org/tensorflow/framework/activations/SoftmaxTest.java
+++ b/tensorflow-framework/src/test/java/org/tensorflow/framework/activations/SoftmaxTest.java
@@ -21,8 +21,6 @@
 import org.tensorflow.types.TFloat32;
 import org.tensorflow.types.TFloat64;
 
-import static org.junit.jupiter.api.Assertions.assertThrows;
-
 /** @author Jim Clarke */
 public class SoftmaxTest {
 
@@ -42,7 +40,6 @@ public void setUp() {}
   @AfterEach
   public void tearDown() {}
 
-
   /** Test of Softmax method, of class Activations. */
   @Test
   public void testSoftmaxOpsOperandFloat() {
@@ -54,7 +51,7 @@ public void testSoftmaxOpsOperandFloat() {
     for (TestSession.Mode tfMode : tfModes)
       try (TestSession session = TestSession.createTestSession(tfMode)) {
         Ops tf = session.getTF();
-        Softmax<TFloat32> instance = new Softmax<>(tf);
+        Softmax instance = new Softmax(tf);
         Operand<TFloat32> result = instance.call(tf.constant(input));
         session.evaluate(tf.constant(expected), result);
       }
@@ -71,7 +68,7 @@ public void testSoftmaxOpsOperandDouble() {
     for (TestSession.Mode tfMode : tfModes)
       try (TestSession session = TestSession.createTestSession(tfMode)) {
         Ops tf = session.getTF();
-        Softmax<TFloat64> instance = new Softmax<>(tf);
+        Softmax instance = new Softmax(tf);
         Operand<TFloat64> result = instance.call(tf.constant(input));
         session.evaluate(tf.constant(expected), result);
       }
@@ -88,7 +85,7 @@ public void testSoftmaxOpsOperandDoubleNegative() {
     for (TestSession.Mode tfMode : tfModes)
       try (TestSession session = TestSession.createTestSession(tfMode)) {
         Ops tf = session.getTF();
-        Softmax<TFloat64> instance = new Softmax<>(tf);
+        Softmax instance = new Softmax(tf);
         Operand<TFloat64> result = instance.call(tf.constant(input));
         session.evaluate(tf.constant(expected), result);
       }
@@ -99,13 +96,13 @@ public void testSoftmaxOpsOperandDoubleNegative() {
   public void testSoftmax1D() {
     double[] input = {1, -2, 3, -4, -5, 6, 7, 8};
     double[] expected = {
-            6.0352829e-04, 3.0047902e-05, 4.4595040e-03, 4.0665414e-06,
-            1.4959969e-06, 8.9571528e-02, 2.4348068e-01, 6.6184908e-01
+      6.0352829e-04, 3.0047902e-05, 4.4595040e-03, 4.0665414e-06,
+      1.4959969e-06, 8.9571528e-02, 2.4348068e-01, 6.6184908e-01
     };
     for (TestSession.Mode tfMode : tfModes)
       try (TestSession session = TestSession.createTestSession(tfMode)) {
         Ops tf = session.getTF();
-        Softmax<TFloat64> instance = new Softmax<>(tf);
+        Softmax instance = new Softmax(tf);
         Operand<TFloat64> result = instance.call(tf.constant(input));
         session.evaluate(tf.constant(expected), result);
       }
@@ -116,13 +113,13 @@ public void testSoftmax1D() {
   public void testSoftmax3D() {
     double[][][] input = {{{1, -2}, {3, -4}}, {{-5, 6}, {-7, 8}}};
     double[][][] expected = {
-            {{9.5257413e-01, 4.7425874e-02}, {9.9908900e-01, 9.1105123e-04}},
-            {{1.6701422e-05, 9.9998331e-01}, {3.0590220e-07, 9.9999964e-01}}
+      {{9.5257413e-01, 4.7425874e-02}, {9.9908900e-01, 9.1105123e-04}},
+      {{1.6701422e-05, 9.9998331e-01}, {3.0590220e-07, 9.9999964e-01}}
     };
     for (TestSession.Mode tfMode : tfModes)
       try (TestSession session = TestSession.createTestSession(tfMode)) {
         Ops tf = session.getTF();
-        Softmax<TFloat64> instance = new Softmax<>(tf);
+        Softmax instance = new Softmax(tf);
         Operand<TFloat64> result = instance.call(tf.constant(input));
         session.evaluate(tf.constant(expected), result);
       }
diff --git a/tensorflow-framework/src/test/java/org/tensorflow/framework/activations/SoftplusTest.java b/tensorflow-framework/src/test/java/org/tensorflow/framework/activations/SoftplusTest.java
index a17f2650d62..58a7ddcd3a8 100644
--- a/tensorflow-framework/src/test/java/org/tensorflow/framework/activations/SoftplusTest.java
+++ b/tensorflow-framework/src/test/java/org/tensorflow/framework/activations/SoftplusTest.java
@@ -50,7 +50,7 @@ public void testCallFloat() {
     for (TestSession.Mode tfMode : tfModes)
       try (TestSession session = TestSession.createTestSession(tfMode)) {
         Ops tf = session.getTF();
-        Softplus<TFloat32> instance = new Softplus<>(tf);
+        Softplus instance = new Softplus(tf);
         Operand<TFloat32> result = instance.call(tf.constant(input));
         session.evaluate(expected, result);
       }
@@ -68,7 +68,7 @@ public void testCallDouble() {
     for (TestSession.Mode tfMode : tfModes)
       try (TestSession session = TestSession.createTestSession(tfMode)) {
         Ops tf = session.getTF();
-        Softplus<TFloat64> instance = new Softplus<>(tf);
+        Softplus instance = new Softplus(tf);
         Operand<TFloat64> result = instance.call(tf.constant(input));
         session.evaluate(expected, result);
       }
diff --git a/tensorflow-framework/src/test/java/org/tensorflow/framework/activations/SoftsignTest.java b/tensorflow-framework/src/test/java/org/tensorflow/framework/activations/SoftsignTest.java
index 43591ab4761..bfff18fa9f7 100644
--- a/tensorflow-framework/src/test/java/org/tensorflow/framework/activations/SoftsignTest.java
+++ b/tensorflow-framework/src/test/java/org/tensorflow/framework/activations/SoftsignTest.java
@@ -48,7 +48,7 @@ public void testCallFloat() {
     for (TestSession.Mode tfMode : tfModes)
       try (TestSession session = TestSession.createTestSession(tfMode)) {
         Ops tf = session.getTF();
-        Softsign<TFloat32> instance = new Softsign<>(tf);
+        Softsign instance = new Softsign(tf);
         Operand<TFloat32> result = instance.call(tf.constant(input));
         session.evaluate(expected, result);
       }
@@ -71,7 +71,7 @@ public void testCallDouble() {
     for (TestSession.Mode tfMode : tfModes)
       try (TestSession session = TestSession.createTestSession(tfMode)) {
         Ops tf = session.getTF();
-        Softsign<TFloat64> instance = new Softsign<>(tf);
+        Softsign instance = new Softsign(tf);
         Operand<TFloat64> result = instance.call(tf.constant(input));
         session.evaluate(expected, result);
       }
diff --git a/tensorflow-framework/src/test/java/org/tensorflow/framework/activations/SwishTest.java b/tensorflow-framework/src/test/java/org/tensorflow/framework/activations/SwishTest.java
index 7576789320b..d9add75d8e0 100644
--- a/tensorflow-framework/src/test/java/org/tensorflow/framework/activations/SwishTest.java
+++ b/tensorflow-framework/src/test/java/org/tensorflow/framework/activations/SwishTest.java
@@ -21,8 +21,6 @@
 import org.tensorflow.types.TFloat32;
 import org.tensorflow.types.TFloat64;
 
-import static org.junit.jupiter.api.Assertions.assertThrows;
-
 /** @author Jim Clarke */
 public class SwishTest {
   private final TestSession.Mode[] tfModes = {TestSession.Mode.EAGER, TestSession.Mode.GRAPH};
@@ -41,8 +39,6 @@ public void setUp() {}
   @AfterEach
   public void tearDown() {}
 
-
-
   /** Test of Swish call method */
   @Test
   public void testCallFloat() {
@@ -60,7 +56,7 @@ public void testCallFloat() {
     for (TestSession.Mode tfMode : tfModes)
       try (TestSession session = TestSession.createTestSession(tfMode)) {
         Ops tf = session.getTF();
-        Swish<TFloat32> instance = new Swish<>(tf);
+        Swish instance = new Swish(tf);
         Operand<TFloat32> result = instance.call(tf.constant(input));
         session.evaluate(expected, result);
       }
@@ -83,7 +79,7 @@ public void testCallDouble() {
     for (TestSession.Mode tfMode : tfModes)
       try (TestSession session = TestSession.createTestSession(tfMode)) {
         Ops tf = session.getTF();
-        Swish<TFloat64> instance = new Swish<>(tf);
+        Swish instance = new Swish(tf);
         Operand<TFloat64> result = instance.call(tf.constant(input));
         session.evaluate(expected, result);
       }
diff --git a/tensorflow-framework/src/test/java/org/tensorflow/framework/activations/TanhTest.java b/tensorflow-framework/src/test/java/org/tensorflow/framework/activations/TanhTest.java
index 5162e141c44..98104d552c7 100644
--- a/tensorflow-framework/src/test/java/org/tensorflow/framework/activations/TanhTest.java
+++ b/tensorflow-framework/src/test/java/org/tensorflow/framework/activations/TanhTest.java
@@ -52,7 +52,7 @@ public void testCallFloat() {
     for (TestSession.Mode tfMode : tfModes)
       try (TestSession session = TestSession.createTestSession(tfMode)) {
         Ops tf = session.getTF();
-        Tanh<TFloat32> instance = new Tanh<>(tf);
+        Tanh instance = new Tanh(tf);
         Operand<TFloat32> result = instance.call(tf.constant(input));
         session.evaluate(expected, result);
       }
@@ -71,7 +71,7 @@ public void testCallDouble() {
     for (TestSession.Mode tfMode : tfModes)
       try (TestSession session = TestSession.createTestSession(tfMode)) {
         Ops tf = session.getTF();
-        Tanh<TFloat64> instance = new Tanh<>(tf);
+        Tanh instance = new Tanh(tf);
         Operand<TFloat64> result = instance.call(tf.constant(input));
         session.evaluate(expected, result);
       }
diff --git a/tensorflow-framework/src/test/java/org/tensorflow/framework/initializers/ConstantTest.java b/tensorflow-framework/src/test/java/org/tensorflow/framework/initializers/ConstantTest.java
index 4e81e0620e6..812ef4a1ee3 100644
--- a/tensorflow-framework/src/test/java/org/tensorflow/framework/initializers/ConstantTest.java
+++ b/tensorflow-framework/src/test/java/org/tensorflow/framework/initializers/ConstantTest.java
@@ -51,7 +51,7 @@ public void testCallUInt() {
       try (TestSession session = TestSession.createTestSession(tfMode)) {
         Ops tf = session.getTF();
         Shape shape = Shape.of(2, 2);
-        Constant<TUint8> instance = new Constant<>(tf, 0xf);
+        Constant instance = new Constant(tf, 0xf);
         Operand<TUint8> operand = instance.call(tf.constant(shape), TUint8.class);
         session.evaluate(expected, operand);
       }
@@ -67,7 +67,7 @@ public void testCallInt() {
       try (TestSession session = TestSession.createTestSession(tfMode)) {
         Ops tf = session.getTF();
         Shape shape = Shape.of(2, 2);
-        Constant<TInt32> instance = new Constant<>(tf, 0xf);
+        Constant instance = new Constant(tf, 0xf);
         Operand<TInt32> operand = instance.call(tf.constant(shape), TInt32.class);
         session.evaluate(expected, operand);
       }
@@ -83,7 +83,7 @@ public void testCallLong() {
       try (TestSession session = TestSession.createTestSession(tfMode)) {
         Ops tf = session.getTF();
         Shape shape = Shape.of(2, 2);
-        Constant<TInt64> instance = new Constant<>(tf, 0xffL);
+        Constant instance = new Constant(tf, 0xffL);
         Operand<TInt64> operand = instance.call(tf.constant(shape), TInt64.class);
         session.evaluate(expected, operand);
       }
@@ -97,7 +97,7 @@ public void testCallFloat() {
       try (TestSession session = TestSession.createTestSession(tfMode)) {
         Ops tf = session.getTF();
         Shape shape = Shape.of(2, 2);
-        Constant<TFloat32> instance = new Constant<>(tf, 12.F);
+        Constant instance = new Constant(tf, 12.F);
         Operand<TFloat32> operand = instance.call(tf.constant(shape), TFloat32.class);
         session.evaluate(expected, operand);
       }
@@ -112,7 +112,7 @@ public void testCallDouble() {
         Ops tf = session.getTF();
         Shape shape = Shape.of(2, 2);
 
-        Constant<TFloat64> instance = new Constant<>(tf, 11.);
+        Constant instance = new Constant(tf, 11.);
         Operand<TFloat64> operand = instance.call(tf.constant(shape), TFloat64.class);
         session.evaluate(expected, operand);
       }
@@ -129,7 +129,7 @@ public void testCallString() {
               Ops tf = session.getTF();
               Shape shape = Shape.of(2, 2);
 
-              Constant<TString> instance = new Constant<>(tf, 22);
+              Constant instance = new Constant(tf, 22);
               instance.call(tf.constant(shape), TString.class);
               fail("IllegalArgumentException  should have been thrown for TString");
             }
@@ -145,7 +145,7 @@ public void testCallBool() {
         Shape shape = Shape.of(2, 2);
         Boolean[] expected = {true, true, true, true};
 
-        Constant<TBool> instance = new Constant<>(tf, true);
+        Constant instance = new Constant(tf, true);
         Operand<TBool> operand = instance.call(tf.constant(shape), TBool.class);
         session.evaluate(expected, operand);
       }
@@ -158,7 +158,7 @@ public void testReproducible() {
         Ops tf = session.getTF();
         Shape shape = Shape.of(2, 2);
 
-        Constant<TFloat64> instance = new Constant<>(tf, 11.);
+        Constant instance = new Constant(tf, 11.);
         Operand<TFloat64> operand1 = instance.call(tf.constant(shape), TFloat64.class);
         Operand<TFloat64> operand2 = instance.call(tf.constant(shape), TFloat64.class);
         session.evaluate(operand1, operand2);
diff --git a/tensorflow-framework/src/test/java/org/tensorflow/framework/initializers/GlorotTest.java b/tensorflow-framework/src/test/java/org/tensorflow/framework/initializers/GlorotTest.java
index e9769806928..c46206ad358 100644
--- a/tensorflow-framework/src/test/java/org/tensorflow/framework/initializers/GlorotTest.java
+++ b/tensorflow-framework/src/test/java/org/tensorflow/framework/initializers/GlorotTest.java
@@ -51,7 +51,7 @@ public void testCallNormalFloat() {
       try (TestSession session = TestSession.createTestSession(tfMode)) {
         Ops tf = session.getTF();
         Shape shape = Shape.of(2, 2);
-        Glorot<TFloat32> instance = new Glorot<>(tf, Distribution.TRUNCATED_NORMAL, SEED);
+        Glorot instance = new Glorot(tf, Distribution.TRUNCATED_NORMAL, SEED);
 
         Operand<TFloat32> operand = instance.call(tf.constant(shape), TFloat32.class);
         session.evaluate(expected, operand);
@@ -68,7 +68,7 @@ public void testCallNormalDouble() {
         Ops tf = session.getTF();
         Shape shape = Shape.of(2, 2);
 
-        Glorot<TFloat64> instance = new Glorot<>(tf, Distribution.TRUNCATED_NORMAL, SEED);
+        Glorot instance = new Glorot(tf, Distribution.TRUNCATED_NORMAL, SEED);
         Operand<TFloat64> operand = instance.call(tf.constant(shape), TFloat64.class);
         session.evaluate(expected, operand);
       }
@@ -82,7 +82,7 @@ public void testCallUniformFloat() {
       try (TestSession session = TestSession.createTestSession(tfMode)) {
         Ops tf = session.getTF();
         Shape shape = Shape.of(2, 2);
-        Glorot<TFloat32> instance = new Glorot<>(tf, Distribution.UNIFORM, SEED);
+        Glorot instance = new Glorot(tf, Distribution.UNIFORM, SEED);
         Operand<TFloat32> operand = instance.call(tf.constant(shape), TFloat32.class);
         session.evaluate(expected, operand);
       }
@@ -97,7 +97,7 @@ public void testCallUniformDouble() {
       try (TestSession session = TestSession.createTestSession(tfMode)) {
         Ops tf = session.getTF();
         Shape shape = Shape.of(2, 2);
-        Glorot<TFloat64> instance = new Glorot<>(tf, Distribution.UNIFORM, SEED);
+        Glorot instance = new Glorot(tf, Distribution.UNIFORM, SEED);
         Operand<TFloat64> operand = instance.call(tf.constant(shape), TFloat64.class);
         session.evaluate(expected, operand);
       }
@@ -109,7 +109,7 @@ public void testCallNormalReproducible() {
       try (TestSession session = TestSession.createTestSession(tfMode)) {
         Ops tf = session.getTF();
         Shape shape = Shape.of(2, 2);
-        Glorot<TFloat64> instance = new Glorot<>(tf, Distribution.TRUNCATED_NORMAL, SEED);
+        Glorot instance = new Glorot(tf, Distribution.TRUNCATED_NORMAL, SEED);
         Operand<TFloat64> operand1 = instance.call(tf.constant(shape), TFloat64.class);
         Operand<TFloat64> operand2 = instance.call(tf.constant(shape), TFloat64.class);
         session.evaluate(operand1, operand2);
@@ -122,7 +122,7 @@ public void testCallUniformReproducible() {
       try (TestSession session = TestSession.createTestSession(tfMode)) {
         Ops tf = session.getTF();
         Shape shape = Shape.of(2, 2);
-        Glorot<TFloat64> instance = new Glorot<>(tf, Distribution.UNIFORM, SEED);
+        Glorot instance = new Glorot(tf, Distribution.UNIFORM, SEED);
         Operand<TFloat64> operand1 = instance.call(tf.constant(shape), TFloat64.class);
         Operand<TFloat64> operand2 = instance.call(tf.constant(shape), TFloat64.class);
         session.evaluate(operand1, operand2);
@@ -135,8 +135,7 @@ public void testCallNORMALReproducible() {
       try (TestSession session = TestSession.createTestSession(tfMode)) {
         Ops tf = session.getTF();
         Shape shape = Shape.of(2, 2);
-        Glorot<TFloat64> instance =
-            new Glorot<>(tf, Distribution.NORMAL, SEED);
+        Glorot instance = new Glorot(tf, Distribution.NORMAL, SEED);
         Operand<TFloat64> operand1 = instance.call(tf.constant(shape), TFloat64.class);
         Operand<TFloat64> operand2 = instance.call(tf.constant(shape), TFloat64.class);
         session.evaluate(operand1, operand2);
diff --git a/tensorflow-framework/src/test/java/org/tensorflow/framework/initializers/HeTest.java b/tensorflow-framework/src/test/java/org/tensorflow/framework/initializers/HeTest.java
index 8953fa3005e..2c9ef3fc56d 100644
--- a/tensorflow-framework/src/test/java/org/tensorflow/framework/initializers/HeTest.java
+++ b/tensorflow-framework/src/test/java/org/tensorflow/framework/initializers/HeTest.java
@@ -27,7 +27,6 @@
 public class HeTest {
   private static final long SEED = 1000L;
   private final TestSession.Mode[] tfModes = {TestSession.Mode.EAGER, TestSession.Mode.GRAPH};
-  int counter;
 
   public HeTest() {}
 
@@ -51,7 +50,7 @@ public void testCallNormalFloat() {
       try (TestSession session = TestSession.createTestSession(tfMode)) {
         Ops tf = session.getTF();
         Shape shape = Shape.of(2, 2);
-        He<TFloat32> instance = new He<>(tf, Distribution.TRUNCATED_NORMAL, SEED);
+        He instance = new He(tf, Distribution.TRUNCATED_NORMAL, SEED);
         Operand<TFloat32> operand = instance.call(tf.constant(shape), TFloat32.class);
         session.evaluate(expected, operand);
       }
@@ -66,7 +65,7 @@ public void testCallNormalDouble() {
       try (TestSession session = TestSession.createTestSession(tfMode)) {
         Ops tf = session.getTF();
         Shape shape = Shape.of(2, 2);
-        He<TFloat64> instance = new He<>(tf, Distribution.TRUNCATED_NORMAL, SEED);
+        He instance = new He(tf, Distribution.TRUNCATED_NORMAL, SEED);
         Operand<TFloat64> operand = instance.call(tf.constant(shape), TFloat64.class);
         session.evaluate(expected, operand);
       }
@@ -80,7 +79,7 @@ public void testCallFloat() {
       try (TestSession session = TestSession.createTestSession(tfMode)) {
         Ops tf = session.getTF();
         Shape shape = Shape.of(2, 2);
-        He<TFloat32> instance = new He<>(tf, Distribution.UNIFORM, SEED);
+        He instance = new He(tf, Distribution.UNIFORM, SEED);
         Operand<TFloat32> operand = instance.call(tf.constant(shape), TFloat32.class);
         session.evaluate(expected, operand);
       }
@@ -95,7 +94,7 @@ public void testCallDouble() {
       try (TestSession session = TestSession.createTestSession(tfMode)) {
         Ops tf = session.getTF();
         Shape shape = Shape.of(2, 2);
-        He<TFloat64> instance = new He<>(tf, Distribution.UNIFORM, SEED);
+        He instance = new He(tf, Distribution.UNIFORM, SEED);
         Operand<TFloat64> operand = instance.call(tf.constant(shape), TFloat64.class);
         session.evaluate(expected, operand);
       }
@@ -107,7 +106,7 @@ public void testCallNormalReproducible() {
       try (TestSession session = TestSession.createTestSession(tfMode)) {
         Ops tf = session.getTF();
         Shape shape = Shape.of(2, 2);
-        He<TFloat64> instance = new He<>(tf, Distribution.TRUNCATED_NORMAL, SEED);
+        He instance = new He(tf, Distribution.TRUNCATED_NORMAL, SEED);
         Operand<TFloat64> operand1 = instance.call(tf.constant(shape), TFloat64.class);
         Operand<TFloat64> operand2 = instance.call(tf.constant(shape), TFloat64.class);
         session.evaluate(operand1, operand2);
@@ -120,7 +119,7 @@ public void testCallUniformReproducible() {
       try (TestSession session = TestSession.createTestSession(tfMode)) {
         Ops tf = session.getTF();
         Shape shape = Shape.of(2, 2);
-        He<TFloat64> instance = new He<>(tf, Distribution.UNIFORM, SEED);
+        He instance = new He(tf, Distribution.UNIFORM, SEED);
         Operand<TFloat64> operand1 = instance.call(tf.constant(shape), TFloat64.class);
         Operand<TFloat64> operand2 = instance.call(tf.constant(shape), TFloat64.class);
         session.evaluate(operand1, operand2);
@@ -133,7 +132,7 @@ public void testCallNORMALReproducible() {
       try (TestSession session = TestSession.createTestSession(tfMode)) {
         Ops tf = session.getTF();
         Shape shape = Shape.of(2, 2);
-        He<TFloat64> instance = new He<>(tf, Distribution.NORMAL, SEED);
+        He instance = new He(tf, Distribution.NORMAL, SEED);
         Operand<TFloat64> operand1 = instance.call(tf.constant(shape), TFloat64.class);
         Operand<TFloat64> operand2 = instance.call(tf.constant(shape), TFloat64.class);
         session.evaluate(operand1, operand2);
diff --git a/tensorflow-framework/src/test/java/org/tensorflow/framework/initializers/IdentityTest.java b/tensorflow-framework/src/test/java/org/tensorflow/framework/initializers/IdentityTest.java
index 6eee5473937..ef72422474d 100644
--- a/tensorflow-framework/src/test/java/org/tensorflow/framework/initializers/IdentityTest.java
+++ b/tensorflow-framework/src/test/java/org/tensorflow/framework/initializers/IdentityTest.java
@@ -21,10 +21,6 @@
 import org.tensorflow.op.Ops;
 import org.tensorflow.types.TFloat32;
 import org.tensorflow.types.TFloat64;
-import org.tensorflow.types.TInt32;
-
-import static org.junit.jupiter.api.Assertions.assertThrows;
-import static org.junit.jupiter.api.Assertions.fail;
 
 /** Test the Identity initializer */
 public class IdentityTest {
@@ -64,7 +60,7 @@ public void testCallFloat() {
       try (TestSession session = TestSession.createTestSession(tfMode)) {
         Ops tf = session.getTF();
         Shape shape = Shape.of(10, 10);
-        Identity<TFloat32> instance = new Identity<>(tf, 2.);
+        Identity instance = new Identity(tf, 2.);
         Operand<TFloat32> operand = instance.call(tf.constant(shape), TFloat32.class);
         session.evaluate(expected, operand);
       }
@@ -90,7 +86,7 @@ public void testCallDouble() {
       try (TestSession session = TestSession.createTestSession(tfMode)) {
         Ops tf = session.getTF();
         Shape shape = Shape.of(10, 10);
-        Identity<TFloat64> instance = new Identity<>(tf, 2.);
+        Identity instance = new Identity(tf, 2.);
         Operand<TFloat64> operand = instance.call(tf.constant(shape), TFloat64.class);
         session.evaluate(expected, operand);
       }
@@ -103,7 +99,7 @@ public void testReproducible() {
         Ops tf = session.getTF();
         Shape shape = Shape.of(2, 2);
 
-        Identity<TFloat64> instance = new Identity<>(tf, 2.);
+        Identity instance = new Identity(tf, 2.);
         Operand<TFloat64> operand1 = instance.call(tf.constant(shape), TFloat64.class);
         Operand<TFloat64> operand2 = instance.call(tf.constant(shape), TFloat64.class);
         session.evaluate(operand1, operand2);
diff --git a/tensorflow-framework/src/test/java/org/tensorflow/framework/initializers/LeCunTest.java b/tensorflow-framework/src/test/java/org/tensorflow/framework/initializers/LeCunTest.java
index 336850a5549..b6a5fe2a947 100644
--- a/tensorflow-framework/src/test/java/org/tensorflow/framework/initializers/LeCunTest.java
+++ b/tensorflow-framework/src/test/java/org/tensorflow/framework/initializers/LeCunTest.java
@@ -51,7 +51,7 @@ public void testCallNormalFloat() {
       try (TestSession session = TestSession.createTestSession(tfMode)) {
         Ops tf = session.getTF();
         Shape shape = Shape.of(2, 2);
-        LeCun<TFloat32> instance = new LeCun<>(tf, Distribution.TRUNCATED_NORMAL, SEED);
+        LeCun instance = new LeCun(tf, Distribution.TRUNCATED_NORMAL, SEED);
         Operand<TFloat32> operand = instance.call(tf.constant(shape), TFloat32.class);
         session.evaluate(expected, operand);
       }
@@ -66,7 +66,7 @@ public void testCallNormalDouble() {
       try (TestSession session = TestSession.createTestSession(tfMode)) {
         Ops tf = session.getTF();
         Shape shape = Shape.of(2, 2);
-        LeCun<TFloat64> instance = new LeCun<>(tf, Distribution.TRUNCATED_NORMAL, SEED);
+        LeCun instance = new LeCun(tf, Distribution.TRUNCATED_NORMAL, SEED);
         Operand<TFloat64> operand = instance.call(tf.constant(shape), TFloat64.class);
         session.evaluate(expected, operand);
       }
@@ -80,7 +80,7 @@ public void testCallUniformFloat() {
       try (TestSession session = TestSession.createTestSession(tfMode)) {
         Ops tf = session.getTF();
         Shape shape = Shape.of(2, 2);
-        LeCun<TFloat32> instance = new LeCun<>(tf, Distribution.UNIFORM, SEED);
+        LeCun instance = new LeCun(tf, Distribution.UNIFORM, SEED);
         Operand<TFloat32> operand = instance.call(tf.constant(shape), TFloat32.class);
         session.evaluate(expected, operand);
       }
@@ -95,7 +95,7 @@ public void testCallUniformDouble() {
       try (TestSession session = TestSession.createTestSession(tfMode)) {
         Ops tf = session.getTF();
         Shape shape = Shape.of(2, 2);
-        LeCun<TFloat64> instance = new LeCun<>(tf, Distribution.UNIFORM, SEED);
+        LeCun instance = new LeCun(tf, Distribution.UNIFORM, SEED);
         Operand<TFloat64> operand = instance.call(tf.constant(shape), TFloat64.class);
         session.evaluate(expected, operand);
       }
@@ -107,7 +107,7 @@ public void testCallNormalReproducible() {
       try (TestSession session = TestSession.createTestSession(tfMode)) {
         Ops tf = session.getTF();
         Shape shape = Shape.of(2, 2);
-        LeCun<TFloat64> instance = new LeCun<>(tf, Distribution.TRUNCATED_NORMAL, SEED);
+        LeCun instance = new LeCun(tf, Distribution.TRUNCATED_NORMAL, SEED);
         Operand<TFloat64> operand1 = instance.call(tf.constant(shape), TFloat64.class);
         Operand<TFloat64> operand2 = instance.call(tf.constant(shape), TFloat64.class);
         session.evaluate(operand1, operand2);
@@ -120,7 +120,7 @@ public void testCallUniformReproducible() {
       try (TestSession session = TestSession.createTestSession(tfMode)) {
         Ops tf = session.getTF();
         Shape shape = Shape.of(2, 2);
-        LeCun<TFloat64> instance = new LeCun<>(tf, Distribution.UNIFORM, SEED);
+        LeCun instance = new LeCun(tf, Distribution.UNIFORM, SEED);
         Operand<TFloat64> operand1 = instance.call(tf.constant(shape), TFloat64.class);
         Operand<TFloat64> operand2 = instance.call(tf.constant(shape), TFloat64.class);
         session.evaluate(operand1, operand2);
@@ -133,7 +133,7 @@ public void testCallNORMALReproducible() {
       try (TestSession session = TestSession.createTestSession(tfMode)) {
         Ops tf = session.getTF();
         Shape shape = Shape.of(2, 2);
-        LeCun<TFloat64> instance = new LeCun<>(tf, Distribution.NORMAL, SEED);
+        LeCun instance = new LeCun(tf, Distribution.NORMAL, SEED);
         Operand<TFloat64> operand1 = instance.call(tf.constant(shape), TFloat64.class);
         Operand<TFloat64> operand2 = instance.call(tf.constant(shape), TFloat64.class);
         session.evaluate(operand1, operand2);
diff --git a/tensorflow-framework/src/test/java/org/tensorflow/framework/initializers/OnesTest.java b/tensorflow-framework/src/test/java/org/tensorflow/framework/initializers/OnesTest.java
index 053ba5dd7ff..d37d1f6eb7f 100644
--- a/tensorflow-framework/src/test/java/org/tensorflow/framework/initializers/OnesTest.java
+++ b/tensorflow-framework/src/test/java/org/tensorflow/framework/initializers/OnesTest.java
@@ -51,7 +51,7 @@ public void testCallUInt() {
       try (TestSession session = TestSession.createTestSession(tfMode)) {
         Ops tf = session.getTF();
         Shape shape = Shape.of(2, 2);
-        Ones<TUint8> instance = new Ones<>(tf);
+        Ones instance = new Ones(tf);
         Operand<TUint8> operand = instance.call(tf.constant(shape), TUint8.class);
         session.evaluate(expected, operand);
       }
@@ -65,7 +65,7 @@ public void testCallInt() {
       try (TestSession session = TestSession.createTestSession(tfMode)) {
         Ops tf = session.getTF();
         Shape shape = Shape.of(2, 2);
-        Ones<TInt32> instance = new Ones<>(tf);
+        Ones instance = new Ones(tf);
         Operand<TInt32> operand = instance.call(tf.constant(shape), TInt32.class);
         session.evaluate(expected, operand);
       }
@@ -79,7 +79,7 @@ public void testCallLong() {
       try (TestSession session = TestSession.createTestSession(tfMode)) {
         Ops tf = session.getTF();
         Shape shape = Shape.of(2, 2);
-        Ones<TInt64> instance = new Ones<>(tf);
+        Ones instance = new Ones(tf);
         Operand<TInt64> operand = instance.call(tf.constant(shape), TInt64.class);
         session.evaluate(expected, operand);
       }
@@ -93,7 +93,7 @@ public void testCallFloat() {
       try (TestSession session = TestSession.createTestSession(tfMode)) {
         Ops tf = session.getTF();
         Shape shape = Shape.of(2, 2);
-        Ones<TFloat32> instance = new Ones<>(tf);
+        Ones instance = new Ones(tf);
         Operand<TFloat32> operand = instance.call(tf.constant(shape), TFloat32.class);
         session.evaluate(expected, operand);
       }
@@ -108,7 +108,7 @@ public void testCallDouble() {
         Ops tf = session.getTF();
         Shape shape = Shape.of(2, 2);
 
-        Ones<TFloat64> instance = new Ones<>(tf);
+        Ones instance = new Ones(tf);
         Operand<TFloat64> operand = instance.call(tf.constant(shape), TFloat64.class);
         session.evaluate(expected, operand);
       }
@@ -125,7 +125,7 @@ public void testCallString() {
               Ops tf = session.getTF();
               Shape shape = Shape.of(2, 2);
 
-              Ones<TString> instance = new Ones<>(tf);
+              Ones instance = new Ones(tf);
               instance.call(tf.constant(shape), TString.class);
               fail("IllegalArgumentException should have been thrown for TString");
             }
@@ -140,7 +140,7 @@ public void testCallBool() {
         Ops tf = session.getTF();
         Shape shape = Shape.of(2, 2);
 
-        Ones<TBool> instance = new Ones<>(tf);
+        Ones instance = new Ones(tf);
         Operand<TBool> operand = instance.call(tf.constant(shape), TBool.class);
         session.evaluate(expected, operand);
       }
@@ -153,7 +153,7 @@ public void testReproducible() {
         Ops tf = session.getTF();
         Shape shape = Shape.of(2, 2);
 
-        Ones<TFloat64> instance = new Ones<>(tf);
+        Ones instance = new Ones(tf);
         Operand<TFloat64> operand1 = instance.call(tf.constant(shape), TFloat64.class);
         Operand<TFloat64> operand2 = instance.call(tf.constant(shape), TFloat64.class);
         session.evaluate(operand1, operand2);
diff --git a/tensorflow-framework/src/test/java/org/tensorflow/framework/initializers/OrthogonalTest.java b/tensorflow-framework/src/test/java/org/tensorflow/framework/initializers/OrthogonalTest.java
index 22b89d9177c..0badd39db30 100644
--- a/tensorflow-framework/src/test/java/org/tensorflow/framework/initializers/OrthogonalTest.java
+++ b/tensorflow-framework/src/test/java/org/tensorflow/framework/initializers/OrthogonalTest.java
@@ -21,10 +21,6 @@
 import org.tensorflow.op.Ops;
 import org.tensorflow.types.TFloat32;
 import org.tensorflow.types.TFloat64;
-import org.tensorflow.types.TInt32;
-
-import static org.junit.jupiter.api.Assertions.assertThrows;
-import static org.junit.jupiter.api.Assertions.fail;
 
 /** Test the Orthogonal initializer */
 public class OrthogonalTest {
@@ -156,7 +152,7 @@ public void testCallFloat() {
       try (TestSession session = TestSession.createTestSession(tfMode)) {
         Ops tf = session.getTF();
         Shape shape = Shape.of(10, 10);
-        Orthogonal<TFloat32> instance = new Orthogonal<>(tf, GAIN_VALUE, SEED);
+        Orthogonal instance = new Orthogonal(tf, GAIN_VALUE, SEED);
         Operand<TFloat32> operand = instance.call(tf.constant(shape), TFloat32.class);
         session.evaluate(expected, operand);
       }
@@ -271,7 +267,7 @@ public void testCallDouble() {
       try (TestSession session = TestSession.createTestSession(tfMode)) {
         Ops tf = session.getTF();
         Shape shape = Shape.of(10, 10);
-        Orthogonal<TFloat64> instance = new Orthogonal<>(tf, GAIN_VALUE, SEED);
+        Orthogonal instance = new Orthogonal(tf, GAIN_VALUE, SEED);
         Operand<TFloat64> operand = instance.call(tf.constant(shape), TFloat64.class);
         session.evaluate(expected, operand);
       }
@@ -284,7 +280,7 @@ public void testReproducible() {
         Ops tf = session.getTF();
         Shape shape = Shape.of(2, 2);
 
-        Orthogonal<TFloat64> instance = new Orthogonal<>(tf, GAIN_VALUE, SEED);
+        Orthogonal instance = new Orthogonal(tf, GAIN_VALUE, SEED);
         Operand<TFloat64> operand1 = instance.call(tf.constant(shape), TFloat64.class);
         Operand<TFloat64> operand2 = instance.call(tf.constant(shape), TFloat64.class);
         session.evaluate(operand1, operand2);
diff --git a/tensorflow-framework/src/test/java/org/tensorflow/framework/initializers/RandomNormalTest.java b/tensorflow-framework/src/test/java/org/tensorflow/framework/initializers/RandomNormalTest.java
index 3b2b3bdb243..7f9e5d841f0 100644
--- a/tensorflow-framework/src/test/java/org/tensorflow/framework/initializers/RandomNormalTest.java
+++ b/tensorflow-framework/src/test/java/org/tensorflow/framework/initializers/RandomNormalTest.java
@@ -52,8 +52,7 @@ public void testCalltestSoftmaxFloat() {
       try (TestSession session = TestSession.createTestSession(tfMode)) {
         Ops tf = session.getTF();
         Shape shape = Shape.of(2, 2);
-        RandomNormal<TFloat32> instance =
-            new RandomNormal<>(tf, MEAN_VALUE, STDDEV_VALUE, SEED);
+        RandomNormal instance = new RandomNormal(tf, MEAN_VALUE, STDDEV_VALUE, SEED);
         Operand<TFloat32> operand = instance.call(tf.constant(shape), TFloat32.class);
         session.evaluate(expected, operand);
       }
@@ -68,8 +67,7 @@ public void testCalltestSoftmaxDouble() {
       try (TestSession session = TestSession.createTestSession(tfMode)) {
         Ops tf = session.getTF();
         Shape shape = Shape.of(2, 2);
-        RandomNormal<TFloat64> instance =
-            new RandomNormal<>(tf, MEAN_VALUE, STDDEV_VALUE, SEED);
+        RandomNormal instance = new RandomNormal(tf, MEAN_VALUE, STDDEV_VALUE, SEED);
         Operand<TFloat64> operand = instance.call(tf.constant(shape), TFloat64.class);
         session.evaluate(expected, operand);
       }
@@ -82,8 +80,7 @@ public void testReproducible() {
         Ops tf = session.getTF();
         Shape shape = Shape.of(2, 2);
 
-        RandomNormal<TFloat64> instance =
-            new RandomNormal<>(tf, MEAN_VALUE, STDDEV_VALUE, SEED);
+        RandomNormal instance = new RandomNormal(tf, MEAN_VALUE, STDDEV_VALUE, SEED);
         Operand<TFloat64> operand1 = instance.call(tf.constant(shape), TFloat64.class);
         Operand<TFloat64> operand2 = instance.call(tf.constant(shape), TFloat64.class);
         session.evaluate(operand1, operand2);
diff --git a/tensorflow-framework/src/test/java/org/tensorflow/framework/initializers/RandomUniformTest.java b/tensorflow-framework/src/test/java/org/tensorflow/framework/initializers/RandomUniformTest.java
index 23e26083a9b..9ad6509d40c 100644
--- a/tensorflow-framework/src/test/java/org/tensorflow/framework/initializers/RandomUniformTest.java
+++ b/tensorflow-framework/src/test/java/org/tensorflow/framework/initializers/RandomUniformTest.java
@@ -53,8 +53,7 @@ public void testCallInt() {
       try (TestSession session = TestSession.createTestSession(tfMode)) {
         Ops tf = session.getTF();
         Shape shape = Shape.of(2, 2);
-        RandomUniform<TInt32> instance =
-            new RandomUniform<>(tf, MIN_VALUE, MAX_VALUE, SEED);
+        RandomUniform instance = new RandomUniform(tf, MIN_VALUE, MAX_VALUE, SEED);
         Operand<TInt32> operand = instance.call(tf.constant(shape), TInt32.class);
         session.evaluate(expected, operand);
       }
@@ -68,8 +67,7 @@ public void testCallFloat() {
       try (TestSession session = TestSession.createTestSession(tfMode)) {
         Ops tf = session.getTF();
         Shape shape = Shape.of(2, 2);
-        RandomUniform<TFloat32> instance =
-            new RandomUniform<>(tf, MIN_VALUE, MAX_VALUE, SEED);
+        RandomUniform instance = new RandomUniform(tf, MIN_VALUE, MAX_VALUE, SEED);
         Operand<TFloat32> operand = instance.call(tf.constant(shape), TFloat32.class);
         session.evaluate(expected, operand);
       }
@@ -84,8 +82,7 @@ public void testCallDouble() {
       try (TestSession session = TestSession.createTestSession(tfMode)) {
         Ops tf = session.getTF();
         Shape shape = Shape.of(2, 2);
-        RandomUniform<TFloat64> instance =
-            new RandomUniform<>(tf, MIN_VALUE, MAX_VALUE, SEED);
+        RandomUniform instance = new RandomUniform(tf, MIN_VALUE, MAX_VALUE, SEED);
         Operand<TFloat64> operand = instance.call(tf.constant(shape), TFloat64.class);
         session.evaluate(expected, operand);
       }
@@ -98,8 +95,7 @@ public void testReproducible() {
         Ops tf = session.getTF();
         Shape shape = Shape.of(2, 2);
 
-        RandomUniform<TFloat64> instance =
-            new RandomUniform<>(tf, MIN_VALUE, MAX_VALUE, SEED);
+        RandomUniform instance = new RandomUniform(tf, MIN_VALUE, MAX_VALUE, SEED);
         Operand<TFloat64> operand1 = instance.call(tf.constant(shape), TFloat64.class);
         Operand<TFloat64> operand2 = instance.call(tf.constant(shape), TFloat64.class);
         session.evaluate(operand1, operand2);
diff --git a/tensorflow-framework/src/test/java/org/tensorflow/framework/initializers/TruncatedNormalTest.java b/tensorflow-framework/src/test/java/org/tensorflow/framework/initializers/TruncatedNormalTest.java
index 96bf915e199..898cede46b2 100644
--- a/tensorflow-framework/src/test/java/org/tensorflow/framework/initializers/TruncatedNormalTest.java
+++ b/tensorflow-framework/src/test/java/org/tensorflow/framework/initializers/TruncatedNormalTest.java
@@ -52,8 +52,7 @@ public void testCallFloat() {
       try (TestSession session = TestSession.createTestSession(tfMode)) {
         Ops tf = session.getTF();
         Shape shape = Shape.of(2, 2);
-        TruncatedNormal<TFloat32> instance =
-            new TruncatedNormal<>(tf, MEAN_VALUE, STDDEV_VALUE, SEED);
+        TruncatedNormal instance = new TruncatedNormal(tf, MEAN_VALUE, STDDEV_VALUE, SEED);
         Operand<TFloat32> operand = instance.call(tf.constant(shape), TFloat32.class);
         session.evaluate(expected, operand);
       }
@@ -68,8 +67,7 @@ public void testCallDouble() {
       try (TestSession session = TestSession.createTestSession(tfMode)) {
         Ops tf = session.getTF();
         Shape shape = Shape.of(2, 2);
-        TruncatedNormal<TFloat64> instance =
-            new TruncatedNormal<>(tf, MEAN_VALUE, STDDEV_VALUE, SEED);
+        TruncatedNormal instance = new TruncatedNormal(tf, MEAN_VALUE, STDDEV_VALUE, SEED);
         Operand<TFloat64> operand = instance.call(tf.constant(shape), TFloat64.class);
         session.evaluate(expected, operand);
       }
@@ -82,8 +80,7 @@ public void testReproducible() {
         Ops tf = session.getTF();
         Shape shape = Shape.of(2, 2);
 
-        TruncatedNormal<TFloat64> instance =
-            new TruncatedNormal<>(tf, MEAN_VALUE, STDDEV_VALUE, SEED);
+        TruncatedNormal instance = new TruncatedNormal(tf, MEAN_VALUE, STDDEV_VALUE, SEED);
         Operand<TFloat64> operand1 = instance.call(tf.constant(shape), TFloat64.class);
         Operand<TFloat64> operand2 = instance.call(tf.constant(shape), TFloat64.class);
         session.evaluate(operand1, operand2);
diff --git a/tensorflow-framework/src/test/java/org/tensorflow/framework/initializers/VarianceScalingTest.java b/tensorflow-framework/src/test/java/org/tensorflow/framework/initializers/VarianceScalingTest.java
index 159affb07e2..5a2819e64c0 100644
--- a/tensorflow-framework/src/test/java/org/tensorflow/framework/initializers/VarianceScalingTest.java
+++ b/tensorflow-framework/src/test/java/org/tensorflow/framework/initializers/VarianceScalingTest.java
@@ -50,8 +50,8 @@ public void testCallFloat1FanInTruncatedNormal() {
       try (TestSession session = TestSession.createTestSession(tfMode)) {
         Ops tf = session.getTF();
         Shape shape = Shape.of(2, 2);
-        VarianceScaling<TFloat32> instance =
-            new VarianceScaling<>(
+        VarianceScaling instance =
+            new VarianceScaling(
                 tf,
                 1.0,
                 VarianceScaling.Mode.FAN_IN,
@@ -71,8 +71,8 @@ public void testCallDouble1FanInTruncatedNormal() {
       try (TestSession session = TestSession.createTestSession(tfMode)) {
         Ops tf = session.getTF();
         Shape shape = Shape.of(2, 2);
-        VarianceScaling<TFloat64> instance =
-            new VarianceScaling<>(
+        VarianceScaling instance =
+            new VarianceScaling(
                 tf,
                 1.0,
                 VarianceScaling.Mode.FAN_IN,
@@ -91,13 +91,9 @@ public void testCallFloat1FanInNormal() {
       try (TestSession session = TestSession.createTestSession(tfMode)) {
         Ops tf = session.getTF();
         Shape shape = Shape.of(2, 2);
-        VarianceScaling<TFloat32> instance =
-            new VarianceScaling<>(
-                tf,
-                1.0,
-                VarianceScaling.Mode.FAN_IN,
-                VarianceScaling.Distribution.NORMAL,
-                SEED);
+        VarianceScaling instance =
+            new VarianceScaling(
+                tf, 1.0, VarianceScaling.Mode.FAN_IN, VarianceScaling.Distribution.NORMAL, SEED);
         Operand<TFloat32> operand = instance.call(tf.constant(shape), TFloat32.class);
         session.evaluate(expected, operand);
       }
@@ -112,13 +108,9 @@ public void testCalltestSoftmaxDouble1FanInNormal() {
       try (TestSession session = TestSession.createTestSession(tfMode)) {
         Ops tf = session.getTF();
         Shape shape = Shape.of(2, 2);
-        VarianceScaling<TFloat64> instance =
-            new VarianceScaling<>(
-                tf,
-                1.0,
-                VarianceScaling.Mode.FAN_IN,
-                VarianceScaling.Distribution.NORMAL,
-                SEED);
+        VarianceScaling instance =
+            new VarianceScaling(
+                tf, 1.0, VarianceScaling.Mode.FAN_IN, VarianceScaling.Distribution.NORMAL, SEED);
         Operand<TFloat64> operand = instance.call(tf.constant(shape), TFloat64.class);
         session.evaluate(expected, operand);
       }
@@ -132,8 +124,8 @@ public void testCalltestSoftmaxFloat1FanInUNIFORM() {
       try (TestSession session = TestSession.createTestSession(tfMode)) {
         Ops tf = session.getTF();
         Shape shape = Shape.of(2, 2);
-        VarianceScaling<TFloat32> instance =
-            new VarianceScaling<>(
+        VarianceScaling instance =
+            new VarianceScaling(
                 tf, 1.0, VarianceScaling.Mode.FAN_IN, VarianceScaling.Distribution.UNIFORM, SEED);
         Operand<TFloat32> operand = instance.call(tf.constant(shape), TFloat32.class);
         session.evaluate(expected, operand);
@@ -149,8 +141,8 @@ public void testCalltestSoftmaxDouble1FanInUNIFORM() {
       try (TestSession session = TestSession.createTestSession(tfMode)) {
         Ops tf = session.getTF();
         Shape shape = Shape.of(2, 2);
-        VarianceScaling<TFloat64> instance =
-            new VarianceScaling<>(
+        VarianceScaling instance =
+            new VarianceScaling(
                 tf, 1.0, VarianceScaling.Mode.FAN_IN, VarianceScaling.Distribution.UNIFORM, SEED);
         Operand<TFloat64> operand = instance.call(tf.constant(shape), TFloat64.class);
         session.evaluate(expected, operand);
@@ -164,8 +156,8 @@ public void testReproducible1() {
         Ops tf = session.getTF();
         Shape shape = Shape.of(2, 2);
 
-        VarianceScaling<TFloat64> instance =
-            new VarianceScaling<>(
+        VarianceScaling instance =
+            new VarianceScaling(
                 tf, 1.0, VarianceScaling.Mode.FAN_IN, VarianceScaling.Distribution.UNIFORM, SEED);
         Operand<TFloat64> operand1 = instance.call(tf.constant(shape), TFloat64.class);
         Operand<TFloat64> operand2 = instance.call(tf.constant(shape), TFloat64.class);
@@ -180,13 +172,9 @@ public void testReproducible2() {
         Ops tf = session.getTF();
         Shape shape = Shape.of(2, 2);
 
-        VarianceScaling<TFloat64> instance =
-            new VarianceScaling<>(
-                tf,
-                1.0,
-                VarianceScaling.Mode.FAN_IN,
-                VarianceScaling.Distribution.NORMAL,
-                SEED);
+        VarianceScaling instance =
+            new VarianceScaling(
+                tf, 1.0, VarianceScaling.Mode.FAN_IN, VarianceScaling.Distribution.NORMAL, SEED);
         Operand<TFloat64> operand1 = instance.call(tf.constant(shape), TFloat64.class);
         Operand<TFloat64> operand2 = instance.call(tf.constant(shape), TFloat64.class);
         session.evaluate(operand1, operand2);
@@ -200,8 +188,8 @@ public void testReproducible3() {
         Ops tf = session.getTF();
         Shape shape = Shape.of(2, 2);
 
-        VarianceScaling<TFloat64> instance =
-            new VarianceScaling<>(
+        VarianceScaling instance =
+            new VarianceScaling(
                 tf,
                 1.0,
                 VarianceScaling.Mode.FAN_OUT,
@@ -220,8 +208,8 @@ public void testReproducible4() {
         Ops tf = session.getTF();
         Shape shape = Shape.of(2, 2);
 
-        VarianceScaling<TFloat64> instance =
-            new VarianceScaling<>(
+        VarianceScaling instance =
+            new VarianceScaling(
                 tf, 1.0, VarianceScaling.Mode.FAN_AVG, VarianceScaling.Distribution.UNIFORM, SEED);
         Operand<TFloat64> operand1 = instance.call(tf.constant(shape), TFloat64.class);
         Operand<TFloat64> operand2 = instance.call(tf.constant(shape), TFloat64.class);
diff --git a/tensorflow-framework/src/test/java/org/tensorflow/framework/initializers/ZerosTest.java b/tensorflow-framework/src/test/java/org/tensorflow/framework/initializers/ZerosTest.java
index 21bad6ff360..9e80abb61b7 100644
--- a/tensorflow-framework/src/test/java/org/tensorflow/framework/initializers/ZerosTest.java
+++ b/tensorflow-framework/src/test/java/org/tensorflow/framework/initializers/ZerosTest.java
@@ -48,7 +48,7 @@ public void testCallUInt() {
       try (TestSession session = TestSession.createTestSession(tfMode)) {
         Ops tf = session.getTF();
         Shape shape = Shape.of(2, 2);
-        Zeros<TUint8> instance = new Zeros<>(tf);
+        Zeros instance = new Zeros(tf);
         Operand<TUint8> operand = instance.call(tf.constant(shape), TUint8.class);
         session.evaluate(expected, operand);
       }
@@ -62,7 +62,7 @@ public void testCallInt() {
       try (TestSession session = TestSession.createTestSession(tfMode)) {
         Ops tf = session.getTF();
         Shape shape = Shape.of(2, 2);
-        Zeros<TInt32> instance = new Zeros<>(tf);
+        Zeros instance = new Zeros(tf);
         Operand<TInt32> operand = instance.call(tf.constant(shape), TInt32.class);
         session.evaluate(expected, operand);
       }
@@ -76,7 +76,7 @@ public void testCallLong() {
       try (TestSession session = TestSession.createTestSession(tfMode)) {
         Ops tf = session.getTF();
         Shape shape = Shape.of(2, 2);
-        Zeros<TInt64> instance = new Zeros<>(tf);
+        Zeros instance = new Zeros(tf);
         Operand<TInt64> operand = instance.call(tf.constant(shape), TInt64.class);
         session.evaluate(expected, operand);
       }
@@ -90,7 +90,7 @@ public void testCallFloat() {
       try (TestSession session = TestSession.createTestSession(tfMode)) {
         Ops tf = session.getTF();
         Shape shape = Shape.of(2, 2);
-        Zeros<TFloat32> instance = new Zeros<>(tf);
+        Zeros instance = new Zeros(tf);
         Operand<TFloat32> operand = instance.call(tf.constant(shape), TFloat32.class);
         session.evaluate(expected, operand);
       }
@@ -105,7 +105,7 @@ public void testCallDouble() {
         Ops tf = session.getTF();
         Shape shape = Shape.of(2, 2);
 
-        Zeros<TFloat64> instance = new Zeros<>(tf);
+        Zeros instance = new Zeros(tf);
         Operand<TFloat64> operand = instance.call(tf.constant(shape), TFloat64.class);
         session.evaluate(expected, operand);
       }
@@ -119,7 +119,7 @@ public void testCallString() {
         Ops tf = session.getTF();
         Shape shape = Shape.of(2, 2);
 
-        Zeros<TString> instance = new Zeros<>(tf);
+        Zeros instance = new Zeros(tf);
         Operand<TString> operand = instance.call(tf.constant(shape), TString.class);
         session.evaluateString(operand, String::isEmpty);
       }
@@ -134,7 +134,7 @@ public void testCallBool() {
         Ops tf = session.getTF();
         Shape shape = Shape.of(2, 2);
 
-        Zeros<TBool> instance = new Zeros<>(tf);
+        Zeros instance = new Zeros(tf);
         Operand<TBool> operand = instance.call(tf.constant(shape), TBool.class);
         session.evaluate(expected, operand);
       }
@@ -147,7 +147,7 @@ public void testReproducible() {
         Ops tf = session.getTF();
         Shape shape = Shape.of(2, 2);
 
-        Zeros<TFloat64> instance = new Zeros<>(tf);
+        Zeros instance = new Zeros(tf);
         Operand<TFloat64> operand1 = instance.call(tf.constant(shape), TFloat64.class);
         Operand<TFloat64> operand2 = instance.call(tf.constant(shape), TFloat64.class);
         session.evaluate(operand1, operand2);

From 309b834fdeb2d1fd881598b32217cf1eaf51e008 Mon Sep 17 00:00:00 2001
From: Jim Clarke <JimClarke5@me.com>
Date: Wed, 3 Mar 2021 11:46:38 -0500
Subject: [PATCH 02/13] resynch with master, for some reason when I build on
 mac, the order for the other xxxxOps classes changes.

---
 .../src/gen/annotations/org/tensorflow/op/Ops.java   | 12 ++++++------
 1 file changed, 6 insertions(+), 6 deletions(-)

diff --git a/tensorflow-core/tensorflow-core-api/src/gen/annotations/org/tensorflow/op/Ops.java b/tensorflow-core/tensorflow-core-api/src/gen/annotations/org/tensorflow/op/Ops.java
index acbae4dac6b..ea3ef31313e 100644
--- a/tensorflow-core/tensorflow-core-api/src/gen/annotations/org/tensorflow/op/Ops.java
+++ b/tensorflow-core/tensorflow-core-api/src/gen/annotations/org/tensorflow/op/Ops.java
@@ -354,20 +354,20 @@ public final class Ops {
 
   public final SparseOps sparse;
 
-  public final BitwiseOps bitwise;
-
   public final TpuOps tpu;
 
+  public final BitwiseOps bitwise;
+
   public final MathOps math;
 
   public final AudioOps audio;
 
   public final SignalOps signal;
 
-  public final QuantizationOps quantization;
-
   public final TrainOps train;
 
+  public final QuantizationOps quantization;
+
   private final Scope scope;
 
   private Ops(Scope scope) {
@@ -385,13 +385,13 @@ private Ops(Scope scope) {
     random = new RandomOps(this);
     strings = new StringsOps(this);
     sparse = new SparseOps(this);
-    bitwise = new BitwiseOps(this);
     tpu = new TpuOps(this);
+    bitwise = new BitwiseOps(this);
     math = new MathOps(this);
     audio = new AudioOps(this);
     signal = new SignalOps(this);
-    quantization = new QuantizationOps(this);
     train = new TrainOps(this);
+    quantization = new QuantizationOps(this);
   }
 
   /**

From c5d37bf7ee07f43985ecc742a9ae5236abbd7d94 Mon Sep 17 00:00:00 2001
From: Jim Clarke <JimClarke5@me.com>
Date: Thu, 4 Mar 2021 16:11:12 -0500
Subject: [PATCH 03/13] Add GeLU activation present in TF 2.4

---
 .../framework/activations/GeLU.java           | 132 ++++++++++++++++++
 .../framework/activations/GeLUTest.java       |  88 ++++++++++++
 2 files changed, 220 insertions(+)
 create mode 100644 tensorflow-framework/src/main/java/org/tensorflow/framework/activations/GeLU.java
 create mode 100644 tensorflow-framework/src/test/java/org/tensorflow/framework/activations/GeLUTest.java

diff --git a/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/GeLU.java b/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/GeLU.java
new file mode 100644
index 00000000000..fa267114781
--- /dev/null
+++ b/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/GeLU.java
@@ -0,0 +1,132 @@
+/* Copyright 2020 The TensorFlow Authors. 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.
+You may obtain a copy of the License at
+
+    http://www.apache.org/licenses/LICENSE-2.0
+
+Unless required by applicable law or agreed to in writing, software
+distributed under the License 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.
+=======================================================================*/
+package org.tensorflow.framework.activations;
+
+import org.tensorflow.Operand;
+import org.tensorflow.Session;
+import org.tensorflow.op.Ops;
+import org.tensorflow.types.TFloat64;
+import org.tensorflow.types.family.TFloating;
+import org.tensorflow.types.family.TNumber;
+
+import java.util.Arrays;
+
+import static org.tensorflow.framework.utils.CastHelper.cast;
+
+/**
+ * Applies the Gaussian error linear unit (GELU) activation function.
+ *
+ * <p>Gaussian error linear unit (GELU) computes {@code x * P(X <= x)}, where {@code P(X) ~ N(0,
+ * 1)}. The (GELU) nonlinearity weights inputs by their value, rather than gates inputs by their
+ * sign as in ReLU. if <code>approximate</code> is <code>true</code> :
+ *
+ * <pre>
+ *     0.5 * x * (1 + tanh(sqrt(2 / pi) * (x + 0.044715 * x^3)))
+ * </pre>
+ *
+ * <p>or, if <code>approximate</code> is <code>false</code>.
+ *
+ * <pre>
+ *     x * P(X <= x) = 0.5 * x * (1 + erf(x / sqrt(2))),
+ * </pre>
+ *
+ * where <code>P(X) ~ N(0, 1)</code>.
+ *
+ * @see <a href="https://arxiv.org/abs/1606.08415">Hendrycks, Dan and Gimpel, Kevin, 2016-2020,
+ *     Gaussian Error Linear Units (GELUs)</a>
+ */
+// TFloating
+public class GeLU extends Activation {
+
+  private final boolean approximate;
+
+  /**
+   * Creates a e Gaussian error linear unit (GELU) activation, with approximate set to false
+   *
+   * @param tf The TensorFlow ops
+   */
+  public GeLU(Ops tf) {
+    this(tf, false);
+  }
+
+  /**
+   * Creates a e Gaussian error linear unit (GELU) activation
+   *
+   * @param tf The TensorFlow ops
+   * @param approximate indicator whether to enable approximation.
+   */
+  public GeLU(Ops tf, boolean approximate) {
+    super(tf);
+    this.approximate = approximate;
+  }
+
+  /** {@inheritDoc} */
+  @Override
+  public <T extends TNumber> Operand<T> call(Operand<T> input) {
+    if (!TFloating.class.isAssignableFrom(input.type())) {
+      throw new IllegalArgumentException(
+          "Tensor type must be numeric or boolean: " + input.type().getSimpleName());
+    }
+    if (approximate) {
+      /*
+              coeff = math_ops.cast(0.044715, features.dtype)
+              return 0.5 * features * (
+                    1.0 + math_ops.tanh(0.7978845608028654 *
+                              (features + coeff * math_ops.pow(features, 3))))
+      */
+      Operand<T> coeff = cast(tf, tf.constant(0.044715), input.type());
+      Operand<T> point5 = cast(tf, tf.constant(0.5), input.type());
+      Operand<T> one = cast(tf, tf.constant(1.0), input.type());
+
+      return tf.math.mul(
+          point5,
+          tf.math.mul(
+              input,
+              tf.math.add(
+                  one,
+                  tf.math.tanh(
+                      tf.math.mul(
+                          // sqrt(2.0 / PI)
+                          cast(tf, tf.constant(0.7978845608028654), input.type()),
+                          tf.math.add(
+                              input,
+                              tf.math.mul(
+                                  coeff,
+                                  tf.math.pow(input, cast(tf, tf.constant(3), input.type()))) // mul
+                              ) // add
+                          ) // mul
+                      ) // tanh
+                  ) // add
+              ) // mul
+          ); // mul
+
+    } else {
+      /*
+      return 0.5 * features * (1.0 + math_ops.erf(
+        features / math_ops.cast(1.4142135623730951, features.dtype)))
+       */
+      return tf.math.mul(
+          cast(tf, tf.constant(0.5), input.type()),
+          tf.math.mul(
+              input,
+              tf.math.add(
+                  cast(tf, tf.constant(1), input.type()),
+                  tf.math.erf(
+                      tf.math.div(
+                          input, cast(tf, tf.constant(1.4142135623730951), input.type()))))));
+    }
+  }
+
+}
diff --git a/tensorflow-framework/src/test/java/org/tensorflow/framework/activations/GeLUTest.java b/tensorflow-framework/src/test/java/org/tensorflow/framework/activations/GeLUTest.java
new file mode 100644
index 00000000000..bea6549dc3e
--- /dev/null
+++ b/tensorflow-framework/src/test/java/org/tensorflow/framework/activations/GeLUTest.java
@@ -0,0 +1,88 @@
+package org.tensorflow.framework.activations;
+
+import org.junit.jupiter.api.Test;
+import org.tensorflow.Operand;
+import org.tensorflow.framework.utils.TestSession;
+import org.tensorflow.op.Ops;
+import org.tensorflow.types.TFloat32;
+import org.tensorflow.types.TFloat64;
+
+import static org.junit.jupiter.api.Assertions.*;
+
+class GeLUTest {
+
+    private final TestSession.Mode[] tfModes = {TestSession.Mode.EAGER, TestSession.Mode.GRAPH};
+
+    /** Test of GeLU call method */
+    @Test
+    public void testCallFloat() {
+        float[][] input = {
+        {0.22805803f, 0.60407318f, 0.91519962f, 0.35643331f, 0.28702669f},
+        {0.11558246f, 0.57658853f, 0.47569648f, 0.02271072f, 0.24709974f}};
+        float[][] expected = {{0.13459972f, 0.43922312f, 0.75042395f, 0.22784713f, 0.17593417f},
+       {0.06310898f, 0.41392788f, 0.32483157f, 0.01156111f, 0.14766297f}};
+        for (TestSession.Mode tfMode : tfModes)
+            try (TestSession session = TestSession.createTestSession(tfMode)) {
+                Ops tf = session.getTF();
+                GeLU instance = new GeLU(tf);
+                Operand<TFloat32> result = instance.call(tf.constant(input));
+                session.evaluate(tf.constant(expected), result);
+            }
+    }
+
+    /** Test of GeLU call method */
+    @Test
+    public void testCallDouble() {
+        double[][] input = {
+                {0.22805803, 0.60407318, 0.91519962, 0.35643331, 0.28702669},
+                {0.11558246, 0.57658853, 0.47569648, 0.02271072, 0.24709974}};
+        double[][] expected = {
+                {0.13459972, 0.43922312, 0.75042395, 0.22784713, 0.17593417},
+                {0.06310898, 0.41392788, 0.32483157, 0.01156111, 0.14766297}
+        };
+        for (TestSession.Mode tfMode : tfModes)
+            try (TestSession session = TestSession.createTestSession(tfMode)) {
+                Ops tf = session.getTF();
+                GeLU instance = new GeLU(tf);
+                Operand<TFloat64> result = instance.call(tf.constant(input));
+                session.evaluate(tf.constant(expected), result);
+            }
+    }
+
+    /** Test of GeLU call method */
+    @Test
+    public void testCallFloatApproximate() {
+        float[][] input = {
+                {0.22805803f, 0.60407318f, 0.91519962f, 0.35643331f, 0.28702669f},
+                {0.11558246f, 0.57658853f, 0.47569648f, 0.02271072f, 0.24709974f}};
+        float[][] expected = {{0.13459886f, 0.43918941f, 0.75030122f, 0.22784227f, 0.17593207f},
+                {0.06310892f, 0.41389921f, 0.32481722f, 0.01156111f, 0.14766179f}};
+        for (TestSession.Mode tfMode : tfModes)
+            try (TestSession session = TestSession.createTestSession(tfMode)) {
+                Ops tf = session.getTF();
+                GeLU instance = new GeLU(tf, true);
+                Operand<TFloat32> result = instance.call(tf.constant(input));
+                session.evaluate(tf.constant(expected), result);
+            }
+    }
+
+    /** Test of GeLU call method */
+    @Test
+    public void testCallDoubleApproximate() {
+        double[][] input = {
+                {0.22805803, 0.60407318, 0.91519962, 0.35643331, 0.28702669},
+                {0.11558246, 0.57658853, 0.47569648, 0.02271072, 0.24709974}};
+        double[][] expected = {{0.13459886, 0.43918941, 0.75030122, 0.22784227, 0.17593207},
+                {0.06310892, 0.41389921, 0.32481722, 0.01156111, 0.14766179}};
+        //for (TestSession.Mode tfMode : tfModes)
+            try (TestSession session = TestSession.createTestSession(TestSession.Mode.GRAPH)) {
+                Ops tf = session.getTF();
+                GeLU instance = new GeLU(tf, true);
+                Operand<TFloat64> result = instance.call(tf.constant(input));
+                session.print(result);
+                session.evaluate(tf.constant(expected), result);
+            }
+    }
+
+
+}
\ No newline at end of file

From 11f8ac9d5fd55ed022f9159c1cc2ef572b0e2555 Mon Sep 17 00:00:00 2001
From: Jim Clarke <JimClarke5@me.com>
Date: Thu, 4 Mar 2021 16:11:42 -0500
Subject: [PATCH 04/13] Fix @param<T> and reformat

---
 .../gen/annotations/org/tensorflow/op/Ops.java  | 12 ++++++------
 .../framework/activations/Activation.java       |  6 ++----
 .../tensorflow/framework/activations/ELU.java   | 14 ++++++++------
 .../framework/activations/Exponential.java      |  9 ++++++++-
 .../framework/activations/HardSigmoid.java      |  8 +-------
 .../tensorflow/framework/activations/ReLU.java  |  1 +
 .../tensorflow/framework/activations/SELU.java  |  9 ++-------
 .../framework/activations/Sigmoid.java          |  9 ++-------
 .../framework/activations/Softmax.java          |  9 ++-------
 .../framework/activations/Softplus.java         |  8 ++------
 .../framework/activations/Softsign.java         |  9 ++-------
 .../tensorflow/framework/activations/Swish.java |  1 +
 .../tensorflow/framework/activations/Tanh.java  |  1 +
 .../framework/initializers/Constant.java        |  8 +++++---
 .../framework/initializers/Identity.java        |  8 +++++---
 .../framework/initializers/Initializer.java     |  4 ++--
 .../tensorflow/framework/initializers/Ones.java |  5 ++++-
 .../framework/initializers/Orthogonal.java      |  4 ++--
 .../framework/initializers/RandomNormal.java    |  5 +++--
 .../framework/initializers/RandomUniform.java   |  1 +
 .../framework/initializers/TruncatedNormal.java |  1 +
 .../framework/initializers/VarianceScaling.java |  1 +
 .../framework/initializers/Zeros.java           |  1 +
 .../framework/activations/ExponentialTest.java  | 15 +++++++++++++++
 .../framework/activations/HardSigmoidTest.java  | 17 +++++++++++++++++
 25 files changed, 95 insertions(+), 71 deletions(-)

diff --git a/tensorflow-core/tensorflow-core-api/src/gen/annotations/org/tensorflow/op/Ops.java b/tensorflow-core/tensorflow-core-api/src/gen/annotations/org/tensorflow/op/Ops.java
index ea3ef31313e..acbae4dac6b 100644
--- a/tensorflow-core/tensorflow-core-api/src/gen/annotations/org/tensorflow/op/Ops.java
+++ b/tensorflow-core/tensorflow-core-api/src/gen/annotations/org/tensorflow/op/Ops.java
@@ -354,20 +354,20 @@ public final class Ops {
 
   public final SparseOps sparse;
 
-  public final TpuOps tpu;
-
   public final BitwiseOps bitwise;
 
+  public final TpuOps tpu;
+
   public final MathOps math;
 
   public final AudioOps audio;
 
   public final SignalOps signal;
 
-  public final TrainOps train;
-
   public final QuantizationOps quantization;
 
+  public final TrainOps train;
+
   private final Scope scope;
 
   private Ops(Scope scope) {
@@ -385,13 +385,13 @@ private Ops(Scope scope) {
     random = new RandomOps(this);
     strings = new StringsOps(this);
     sparse = new SparseOps(this);
-    tpu = new TpuOps(this);
     bitwise = new BitwiseOps(this);
+    tpu = new TpuOps(this);
     math = new MathOps(this);
     audio = new AudioOps(this);
     signal = new SignalOps(this);
-    train = new TrainOps(this);
     quantization = new QuantizationOps(this);
+    train = new TrainOps(this);
   }
 
   /**
diff --git a/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Activation.java b/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Activation.java
index 104e3726ee4..d3546e99a7b 100644
--- a/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Activation.java
+++ b/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Activation.java
@@ -18,9 +18,7 @@
 import org.tensorflow.op.Ops;
 import org.tensorflow.types.family.TNumber;
 
-/**
- * Abstract base class for Activations
- */
+/** Abstract base class for Activations */
 public abstract class Activation {
 
   /** The TensorFlow Ops */
@@ -57,8 +55,8 @@ protected void setTF(Ops tf) {
    * Gets the calculation operation for the activation.
    *
    * @param input the input tensor
-   * @return The operand for the activation
    * @param <T> the data type of the input and result
+   * @return The operand for the activation
    */
   public abstract <T extends TNumber> Operand<T> call(Operand<T> input);
 }
diff --git a/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/ELU.java b/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/ELU.java
index e7ad39fde62..81b34ca77e3 100644
--- a/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/ELU.java
+++ b/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/ELU.java
@@ -17,6 +17,7 @@
 import org.tensorflow.Operand;
 import org.tensorflow.op.Ops;
 import org.tensorflow.types.TBool;
+import org.tensorflow.types.family.TFloating;
 import org.tensorflow.types.family.TNumber;
 
 /**
@@ -47,6 +48,7 @@
  * @see <a href="https://arxiv.org/abs/1511.07289">Clevert et al, 2016, Fast and Accurate Deep
  *     Network Learning by Exponential Linear Units (ELUs)</a>
  */
+// TFloating
 public class ELU extends Activation {
 
   private static final double ALPHA_DEFAULT = 1.0;
@@ -75,15 +77,15 @@ public ELU(Ops tf, double alpha) {
     this.alpha = alpha;
   }
 
-  /**
-   * Gets the calculation operation for the activation.
-   *
-   * @param input the input tensor
-   * @return The operand for the activation
-   */
+  /** {@inheritDoc} */
   @Override
   public <T extends TNumber> Operand<T> call(Operand<T> input) {
 
+    if (!TFloating.class.isAssignableFrom(input.type()) ) {
+      throw new IllegalArgumentException(
+              "Tensor type must be numeric or boolean: " + input.type().getSimpleName());
+    }
+
     Operand<T> result = tf.nn.elu(input);
     if (alpha == 1.0) {
       return result;
diff --git a/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Exponential.java b/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Exponential.java
index 2debca9ddac..6638b9f2e90 100644
--- a/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Exponential.java
+++ b/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Exponential.java
@@ -16,6 +16,7 @@
 
 import org.tensorflow.Operand;
 import org.tensorflow.op.Ops;
+import org.tensorflow.types.family.TFloating;
 import org.tensorflow.types.family.TNumber;
 
 /**
@@ -31,6 +32,7 @@
  *   // result is [0.04978707f,  0.36787945f,  1.f,  2.7182817f, 20.085537f]
  * </pre>
  */
+// TFloating
 public class Exponential extends Activation {
 
   /**
@@ -46,11 +48,16 @@ public Exponential(Ops tf) {
    * Calculates the Exponential activation.
    *
    * @param input the input tensor
+   * @param <T> the data type of the input and result
    * @return an Operand for the exponential activation: <code>exp(x)</code>.
-   * @param <T> the data type of the activation
    */
   @Override
   public <T extends TNumber> Operand<T> call(Operand<T> input) {
+    if (!TFloating.class.isAssignableFrom(input.type()) ) {
+      throw new IllegalArgumentException(
+              "Tensor type must be numeric or boolean: " + input.type().getSimpleName());
+    }
+
     return tf.math.exp(input);
   }
 }
diff --git a/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/HardSigmoid.java b/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/HardSigmoid.java
index 0e2fd6f9342..bc8f5b55291 100644
--- a/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/HardSigmoid.java
+++ b/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/HardSigmoid.java
@@ -52,13 +52,7 @@ public HardSigmoid(Ops tf) {
     super(tf);
   }
 
-  /**
-   * Gets the calculation operation for the activation.
-   *
-   * @param input the input tensor
-   * @return The operand for the activation
-   * @param <T> the data type of the result
-   */
+  /** {@inheritDoc} */
   @Override
   public <T extends TNumber> Operand<T> call(Operand<T> input) {
     Class<T> inputType = input.type();
diff --git a/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/ReLU.java b/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/ReLU.java
index 974d266143f..b3654093109 100644
--- a/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/ReLU.java
+++ b/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/ReLU.java
@@ -56,6 +56,7 @@
  *     // result is [-0.f, -0.f,  0.f,  0.f, 10.f]
  * </pre>
  */
+// TFloating
 public class ReLU extends Activation {
 
   public static final float ALPHA_DEFAULT = 0.0f;
diff --git a/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/SELU.java b/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/SELU.java
index d3328aeb110..0d94d3c5092 100644
--- a/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/SELU.java
+++ b/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/SELU.java
@@ -44,6 +44,7 @@
  *
  * @see <a href="https://arxiv.org/abs/1706.02515">Klambauer et al., 2017</a>
  */
+// TFloating
 public class SELU extends Activation {
 
   /**
@@ -55,13 +56,7 @@ public SELU(Ops tf) {
     super(tf);
   }
 
-  /**
-   * Gets the calculation operation for the activation.
-   *
-   * @param input the input tensor
-   * @return The operand for the activation
-   * @param <T> the data type of the activation
-   */
+  /** {@inheritDoc} */
   @Override
   public <T extends TNumber> Operand<T> call(Operand<T> input) {
     return tf.nn.selu(input);
diff --git a/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Sigmoid.java b/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Sigmoid.java
index c4781ba50d9..eca6d2c3f8c 100644
--- a/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Sigmoid.java
+++ b/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Sigmoid.java
@@ -39,6 +39,7 @@
  *     //                 5.0000000e-01f,7.3105860e-01f, 1.f]
  * </pre>
  */
+// TFloating
 public class Sigmoid extends Activation {
 
   /**
@@ -50,13 +51,7 @@ public Sigmoid(Ops tf) {
     super(tf);
   }
 
-  /**
-   * Gets the calculation operation for the activation.
-   *
-   * @param input the input tensor
-   * @return The operand for the activation
-   * @param <T> the data type of the activation
-   */
+  /** {@inheritDoc} */
   @Override
   public <T extends TNumber> Operand<T> call(Operand<T> input) {
     return tf.math.sigmoid(input);
diff --git a/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Softmax.java b/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Softmax.java
index 8051cee9c04..6991d7d233c 100644
--- a/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Softmax.java
+++ b/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Softmax.java
@@ -36,6 +36,7 @@
  *
  * <p>The input values in are the log-odds of the resulting probability.
  */
+// TFloating
 public class Softmax extends Activation {
 
   private static final int AXIS_DEFAULT = -1;
@@ -63,13 +64,7 @@ public Softmax(Ops tf, int axis) {
     this.axis = axis;
   }
 
-  /**
-   * Gets the calculation operation for the activation.
-   *
-   * @param input the input tensor
-   * @return The operand for the activation
-   * @param <T> the data type of the activation
-   */
+  /** {@inheritDoc} */
   @Override
   public <T extends TNumber> Operand<T> call(Operand<T> input) {
     Shape shape = input.shape();
diff --git a/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Softplus.java b/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Softplus.java
index 762c0ae6bcd..f34d756afda 100644
--- a/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Softplus.java
+++ b/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Softplus.java
@@ -32,6 +32,7 @@
  *     //                 1.3132616e+00f, 2.0000000e+01f]
  * </pre>
  */
+// TFloating
 public class Softplus extends Activation {
 
   /**
@@ -43,12 +44,7 @@ public Softplus(Ops tf) {
     super(tf);
   }
 
-  /**
-   * Gets the calculation operation for the activation.
-   *
-   * @param input the input tensor
-   * @return The operand for the activation
-   */
+  /** {@inheritDoc} */
   @Override
   public <T extends TNumber> Operand<T> call(Operand<T> input) {
     return tf.math.softplus(input);
diff --git a/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Softsign.java b/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Softsign.java
index b6e9f874914..db308154e2f 100644
--- a/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Softsign.java
+++ b/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Softsign.java
@@ -31,6 +31,7 @@
  *     // result is [-0.5f, 0.f, 0.5f]
  * </pre>
  */
+// TFloating
 public class Softsign extends Activation {
 
   /**
@@ -42,13 +43,7 @@ public Softsign(Ops tf) {
     super(tf);
   }
 
-  /**
-   * Gets the calculation operation for the activation.
-   *
-   * @param input the input tensor
-   * @return The operand for the activation
-   * @param <T> the data type of the activation
-   */
+  /** {@inheritDoc} */
   @Override
   public <T extends TNumber> Operand<T> call(Operand<T> input) {
     return tf.nn.softsign(input);
diff --git a/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Swish.java b/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Swish.java
index 5a6fc4c7765..233e9016b56 100644
--- a/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Swish.java
+++ b/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Swish.java
@@ -39,6 +39,7 @@
  *
  * @see <a href="https://arxiv.org/abs/1710.05941">Ramachandran et al., 2017</a>
  */
+// TFloating
 public class Swish extends Activation {
 
   /**
diff --git a/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Tanh.java b/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Tanh.java
index a485638aa4a..b4d3499ca6f 100644
--- a/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Tanh.java
+++ b/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Tanh.java
@@ -31,6 +31,7 @@
  *     // result = [-0.9950547f, -0.7615942f,  0.f,  0.7615942f,  0.9950547f]
  * </pre>
  */
+// TFloating
 public class Tanh extends Activation {
 
   /**
diff --git a/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Constant.java b/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Constant.java
index 5aae5b90e5b..1cd67c38a7d 100644
--- a/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Constant.java
+++ b/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Constant.java
@@ -21,6 +21,8 @@
 import org.tensorflow.types.family.TNumber;
 import org.tensorflow.types.family.TType;
 
+import static org.tensorflow.framework.utils.CastHelper.cast;
+
 /**
  * Initializer that generates tensors with a constant value.
  *
@@ -91,11 +93,11 @@ public <T extends TType> Operand<T> call(Operand<TInt64> dims, Class<T> type) {
     }
     switch (valueType) {
       case LONG:
-        return tf.fill(dims, tf.dtypes.cast(tf.constant(longValue), type));
+        return tf.fill(dims, cast(tf, tf.constant(longValue), type));
       case DOUBLE:
-        return tf.fill(dims, tf.dtypes.cast(tf.constant(doubleValue), type));
+        return tf.fill(dims, cast(tf, tf.constant(doubleValue), type));
       default:
-        return tf.fill(dims, tf.dtypes.cast(tf.constant(booleanValue), type));
+        return tf.fill(dims, cast(tf, tf.constant(booleanValue), type));
     }
   }
 
diff --git a/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Identity.java b/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Identity.java
index e1f50e145a2..560aa394454 100644
--- a/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Identity.java
+++ b/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Identity.java
@@ -21,6 +21,8 @@
 import org.tensorflow.types.TInt64;
 import org.tensorflow.types.family.TType;
 
+import static org.tensorflow.framework.utils.CastHelper.cast;
+
 /**
  * Initializer that generates the identity matrix.
  *
@@ -73,9 +75,9 @@ public <T extends TType> Operand<T> call(Operand<TInt64> dims, Class<T> type) {
     Shape diagShape = Shape.of(diagSize);
 
     Operand<T> op;
-    Operand<T> zero = tf.dtypes.cast(tf.constant(0), type);
+    Operand<T> zero = cast(tf, tf.constant(0), type);
     Operand<T> diagOnes =
-        tf.fill(tf.constant(diagShape.asArray()), tf.dtypes.cast(tf.constant(1.0), type));
+        tf.fill(tf.constant(diagShape.asArray()), cast(tf, tf.constant(1.0), type));
     if (isSquare) {
       op =
           tf.linalg.matrixDiag(
@@ -89,6 +91,6 @@ public <T extends TType> Operand<T> call(Operand<TInt64> dims, Class<T> type) {
       op = tf.linalg.matrixSetDiag(zeroMatrix, diagOnes, tf.constant(0));
     }
 
-    return tf.math.mul(op, tf.dtypes.cast(tf.constant(gain), type));
+    return tf.math.mul(op, cast(tf, tf.constant(gain), type));
   }
 }
diff --git a/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Initializer.java b/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Initializer.java
index 032f9c92792..c22a2582d35 100644
--- a/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Initializer.java
+++ b/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Initializer.java
@@ -25,9 +25,9 @@ public interface Initializer {
    * Generates the operation used to perform the initialization.
    *
    * @param dims the shape dimensions
-   * @param type the type of tensor
-   * @return An operand for the initialization.
+   * @param type the data type of tensor
    * @param <T> The data Type for initializer operation
+   * @return An operand for the initialization.
    */
   <T extends TType> Operand<T> call(Operand<TInt64> dims, Class<T> type);
 }
diff --git a/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Ones.java b/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Ones.java
index 9094c2add97..52528df784f 100644
--- a/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Ones.java
+++ b/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Ones.java
@@ -21,6 +21,8 @@
 import org.tensorflow.types.family.TNumber;
 import org.tensorflow.types.family.TType;
 
+import static org.tensorflow.framework.utils.CastHelper.cast;
+
 /**
  * Initializer that generates tensors initialized to 1.
  *
@@ -60,6 +62,7 @@ public <T extends TType> Operand<T> call(Operand<TInt64> dims, Class<T> type) {
       throw new IllegalArgumentException(
           "Tensor type must be numeric or boolean: " + type.getSimpleName());
     }
-    return tf.fill(dims, tf.dtypes.cast(tf.constant(1.0), type));
+
+    return cast(tf, tf.fill(dims, tf.constant(1)), type);
   }
 }
diff --git a/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Orthogonal.java b/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Orthogonal.java
index b6dffd2d768..de5fb825e42 100644
--- a/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Orthogonal.java
+++ b/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Orthogonal.java
@@ -107,10 +107,10 @@ public <T extends TType> Operand<T> call(Operand<TInt64> dims, Class<T> type) {
     Output<TNumber> qo = qrOp.q();
     Output<TNumber> ro = qrOp.r();
     Operand<TNumber> diagOp =
-        tf.linalg.matrixDiagPart(ro, tf.constant(0), tf.dtypes.cast(tf.constant(0), op.type()));
+        tf.linalg.matrixDiagPart(ro, tf.constant(0), cast(tf, tf.constant(0), op.type()));
     Operand<TNumber> qop = tf.math.mul(qo, tf.math.sign(diagOp));
     if (numRows < numCols) qop = tf.linalg.transpose(qop, null);
 
-    return cast(tf, tf.math.mul(qop, tf.dtypes.cast(tf.constant(this.gain), op.type())), type);
+    return cast(tf, tf.math.mul(qop, cast(tf, tf.constant(this.gain), op.type())), type);
   }
 }
diff --git a/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/RandomNormal.java b/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/RandomNormal.java
index 8f8edcc56ca..7c5d2f9a17b 100644
--- a/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/RandomNormal.java
+++ b/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/RandomNormal.java
@@ -91,11 +91,12 @@ public <T extends TType> Operand<T> call(Operand<TInt64> dims, Class<T> type) {
       throw new IllegalArgumentException("Tensor type must be numeric: " + type.getSimpleName());
     }
     long[] seeds = {seed, 0};
+    // Suppression is ok because it is guarded by the if statement
     @SuppressWarnings("unchecked")
     Class<TNumber> nType = (Class<TNumber>) type;
     Operand<TNumber> distOp = tf.random.statelessRandomNormal(dims, tf.constant(seeds), nType);
     Operand<TNumber> op =
-        tf.math.mul(distOp, tf.dtypes.cast(tf.constant(this.stddev), distOp.type()));
-    return cast(tf, tf.math.add(op, tf.dtypes.cast(tf.constant(mean), distOp.type())), type);
+        tf.math.mul(distOp, cast(tf, tf.constant(this.stddev), distOp.type()));
+    return cast(tf, tf.math.add(op, cast(tf, tf.constant(mean), distOp.type())), type);
   }
 }
diff --git a/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/RandomUniform.java b/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/RandomUniform.java
index 3c7d394d538..30bbd61bae7 100644
--- a/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/RandomUniform.java
+++ b/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/RandomUniform.java
@@ -80,6 +80,7 @@ public <T extends TType> Operand<T> call(Operand<TInt64> dims, Class<T> type) {
     if (!TNumber.class.isAssignableFrom(type)) {
       throw new IllegalArgumentException("Tensor type must be numeric: " + type.getSimpleName());
     }
+    // Suppression is ok because it is guarded by the if statement
     @SuppressWarnings("unchecked")
     Class<TNumber> nType = (Class<TNumber>) type;
     Operand<TNumber> distOp;
diff --git a/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/TruncatedNormal.java b/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/TruncatedNormal.java
index 7969e6988fa..2e53c2db6f9 100644
--- a/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/TruncatedNormal.java
+++ b/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/TruncatedNormal.java
@@ -78,6 +78,7 @@ public <T extends TType> Operand<T> call(Operand<TInt64> dims, Class<T> type) {
     if (!TNumber.class.isAssignableFrom(type)) {
       throw new IllegalArgumentException("Tensor type must be numeric: " + type.getSimpleName());
     }
+    // Suppression is ok because it is guarded by the if statement
     @SuppressWarnings("unchecked")
     Class<TNumber> nType = (Class<TNumber>) type;
     long[] seeds = {seed, 0};
diff --git a/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/VarianceScaling.java b/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/VarianceScaling.java
index 583d5680c57..d820f04a170 100644
--- a/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/VarianceScaling.java
+++ b/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/VarianceScaling.java
@@ -102,6 +102,7 @@ public <T extends TType> Operand<T> call(Operand<TInt64> dims, Class<T> type) {
     if (!TNumber.class.isAssignableFrom(type)) {
       throw new IllegalArgumentException("Tensor type must be numeric: " + type.getSimpleName());
     }
+    // Suppression is ok because it is guarded by the if statement
     @SuppressWarnings("unchecked")
     Class<TNumber> nType = (Class<TNumber>) type;
     Shape shape = ShapeUtils.toShape(this.tf.scope(), dims);
diff --git a/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Zeros.java b/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Zeros.java
index 1f8c08c7910..8368baad804 100644
--- a/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Zeros.java
+++ b/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Zeros.java
@@ -42,6 +42,7 @@ public Zeros(Ops tf) {
     super(tf);
   }
 
+  /** {@inheritDoc} */
   @Override
   public <T extends TType> Operand<T> call(Operand<TInt64> dims, Class<T> type) {
     return tf.zeros(dims, type);
diff --git a/tensorflow-framework/src/test/java/org/tensorflow/framework/activations/ExponentialTest.java b/tensorflow-framework/src/test/java/org/tensorflow/framework/activations/ExponentialTest.java
index e8758148c21..beb5a1d9c9d 100644
--- a/tensorflow-framework/src/test/java/org/tensorflow/framework/activations/ExponentialTest.java
+++ b/tensorflow-framework/src/test/java/org/tensorflow/framework/activations/ExponentialTest.java
@@ -20,6 +20,7 @@
 import org.tensorflow.op.Ops;
 import org.tensorflow.types.TFloat32;
 import org.tensorflow.types.TFloat64;
+import org.tensorflow.types.TInt32;
 
 /** @author Jim Clarke */
 public class ExponentialTest {
@@ -39,6 +40,20 @@ public void setUp() {}
   @AfterEach
   public void tearDown() {}
 
+  /** Test of Exponential call method. */
+  @Test
+  public void testCallInt32() {
+    int[] input = {1, -2, 3, -4, -1, 2, -3, 4};
+    for (TestSession.Mode tfMode : tfModes)
+      try (TestSession session = TestSession.createTestSession(tfMode)) {
+        Ops tf = session.getTF();
+        Exponential instance = new Exponential(tf);
+        Operand<TInt32> result = instance.call(tf.constant(input));
+        session.print(result);
+        //session.evaluate(expected, result);
+      }
+  }
+
   /** Test of Exponential call method. */
   @Test
   public void testCallFloat() {
diff --git a/tensorflow-framework/src/test/java/org/tensorflow/framework/activations/HardSigmoidTest.java b/tensorflow-framework/src/test/java/org/tensorflow/framework/activations/HardSigmoidTest.java
index 59370c745a3..6e8b00dba3e 100644
--- a/tensorflow-framework/src/test/java/org/tensorflow/framework/activations/HardSigmoidTest.java
+++ b/tensorflow-framework/src/test/java/org/tensorflow/framework/activations/HardSigmoidTest.java
@@ -20,6 +20,7 @@
 import org.tensorflow.op.Ops;
 import org.tensorflow.types.TFloat32;
 import org.tensorflow.types.TFloat64;
+import org.tensorflow.types.TInt32;
 
 /** @author Jim Clarke */
 public class HardSigmoidTest {
@@ -66,4 +67,20 @@ public void testCallDouble() {
         session.evaluate(expected, result);
       }
   }
+
+  /** Test of HardSigmoid call method. */
+  @Test
+  public void testCallInt() {
+    int[] input = {-3, -1, 0, 1, 3};
+
+    for (TestSession.Mode tfMode : tfModes)
+      try (TestSession session = TestSession.createTestSession(tfMode)) {
+        Ops tf = session.getTF();
+        HardSigmoid instance = new HardSigmoid(tf);
+        Operand<TInt32> result = instance.call(tf.constant(input));
+        session.print(result);
+        int[] expected = {0, 0, 0, 0, 0};
+        session.evaluate(expected, result);
+      }
+  }
 }

From 40a95afe1b08bb474c3ad5d5130bf7113762813b Mon Sep 17 00:00:00 2001
From: Jim Clarke <JimClarke5@me.com>
Date: Sat, 6 Mar 2021 18:21:38 -0500
Subject: [PATCH 05/13] Fix JavaDoc to add @param <T>

---
 .../framework/initializers/Constant.java      | 68 +++++--------------
 1 file changed, 16 insertions(+), 52 deletions(-)

diff --git a/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Constant.java b/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Constant.java
index 1cd67c38a7d..f20ff694af8 100644
--- a/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Constant.java
+++ b/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Constant.java
@@ -34,76 +34,40 @@
  *      Operand&lt;TFloat32&gt; values =
  *              initializer.call(tf.constant(Shape.of(2,2)), TFloat32.class);
  * </pre>
+ *
+ * <p>Only scalar values are allowed. The constant value provided must be convertible to the data
+ * type requested when calling the initializer.
  */
-public class Constant extends BaseInitializer {
+public class Constant extends BaseInitializer<TType> {
 
-  private final double doubleValue;
-  private final long longValue;
-  private final boolean booleanValue;
-  private final ValueType valueType;
+  private final Operand<? extends TType> value;
 
   /**
    * Creates an Initializer that generates tensors with a constant value.
    *
    * @param tf the TensorFlow Ops
-   * @param value a long value used for the constant.
+   * @param value the value used for the constant.
    */
-  public Constant(Ops tf, long value) {
+  public Constant(Ops tf, Operand<? extends TType> value) {
     super(tf);
-    longValue = value;
-    doubleValue = 0;
-    booleanValue = false;
-    valueType = ValueType.LONG;
+    this.value = value;
   }
 
   /**
-   * Creates an Initializer that generates tensors with a constant value.
+   * Generates the operation used to perform the initialization.
    *
-   * @param tf the TensorFlow Ops
-   * @param value a double value used for the constant.
+   * @param dims the shape dimensions
+   * @param type the data type of tensor
+   * @param <U> The data Type for initializer operation
+   * @return An operand for the initialization.
+   * @throws IllegalArgumentException if the data type is not a TNumber or TBool
    */
-  public Constant(Ops tf, double value) {
-    super(tf);
-    doubleValue = value;
-    longValue = 0;
-    booleanValue = false;
-    valueType = ValueType.DOUBLE;
-  }
-
-  /**
-   * Creates an Initializer that generates tensors with a constant value.
-   *
-   * @param tf the TensorFlow Ops
-   * @param value a boolean value used for the constant.
-   */
-  public Constant(Ops tf, boolean value) {
-    super(tf);
-    booleanValue = value;
-    doubleValue = 0;
-    longValue = 0;
-    valueType = ValueType.BOOLEAN;
-  }
-
-  /** {@inheritDoc} */
   @Override
-  public <T extends TType> Operand<T> call(Operand<TInt64> dims, Class<T> type) {
+  public <U extends TType> Operand<U> call(Operand<TInt64> dims, Class<U> type) {
     if (!TNumber.class.isAssignableFrom(type) && type != TBool.class) {
       throw new IllegalArgumentException(
           "Tensor type must be numeric or boolean: " + type.getSimpleName());
     }
-    switch (valueType) {
-      case LONG:
-        return tf.fill(dims, cast(tf, tf.constant(longValue), type));
-      case DOUBLE:
-        return tf.fill(dims, cast(tf, tf.constant(doubleValue), type));
-      default:
-        return tf.fill(dims, cast(tf, tf.constant(booleanValue), type));
-    }
-  }
-
-  private enum ValueType {
-    LONG,
-    DOUBLE,
-    BOOLEAN
+    return tf.fill(dims, cast(tf, value, type));
   }
 }

From d0e8de92efcd37d103492dd6a8f5bb8afb3918c0 Mon Sep 17 00:00:00 2001
From: Jim Clarke <JimClarke5@me.com>
Date: Sat, 6 Mar 2021 18:23:49 -0500
Subject: [PATCH 06/13] Refactor to add generic to base class and change
 signature of call method to <U extends T>.

Changed all subclasses to match these signatures.
---
 .../framework/activations/Activation.java     |   6 +-
 .../tensorflow/framework/activations/ELU.java |  17 +-
 .../framework/activations/Exponential.java    |  13 +-
 .../framework/activations/GeLU.java           |  24 +--
 .../framework/activations/HardSigmoid.java    |  12 +-
 .../framework/activations/Linear.java         |   4 +-
 .../framework/activations/ReLU.java           |  15 +-
 .../framework/activations/SELU.java           |   7 +-
 .../framework/activations/Sigmoid.java        |   7 +-
 .../framework/activations/Softmax.java        |  11 +-
 .../framework/activations/Softplus.java       |   7 +-
 .../framework/activations/Softsign.java       |   7 +-
 .../framework/activations/Swish.java          |   7 +-
 .../framework/activations/Tanh.java           |   7 +-
 .../framework/constraints/Constraint.java     |   2 +
 .../initializers/BaseInitializer.java         |   3 +-
 .../framework/initializers/Identity.java      |  14 +-
 .../framework/initializers/Initializer.java   |   6 +-
 .../framework/initializers/Ones.java          |  14 +-
 .../framework/initializers/Orthogonal.java    |  32 ++--
 .../framework/initializers/RandomNormal.java  |  20 +--
 .../framework/initializers/RandomUniform.java |  21 +--
 .../initializers/TruncatedNormal.java         |  24 +--
 .../initializers/VarianceScaling.java         |  31 ++--
 .../framework/initializers/Zeros.java         |  20 ++-
 .../activations/ExponentialTest.java          |  15 --
 .../framework/activations/GeLUTest.java       | 149 +++++++++---------
 .../activations/HardSigmoidTest.java          |   1 -
 .../framework/initializers/ConstantTest.java  |  17 +-
 .../framework/initializers/ZerosTest.java     |  22 ++-
 30 files changed, 245 insertions(+), 290 deletions(-)

diff --git a/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Activation.java b/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Activation.java
index d3546e99a7b..708c7daead6 100644
--- a/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Activation.java
+++ b/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Activation.java
@@ -19,7 +19,7 @@
 import org.tensorflow.types.family.TNumber;
 
 /** Abstract base class for Activations */
-public abstract class Activation {
+public abstract class Activation<T extends TNumber> {
 
   /** The TensorFlow Ops */
   protected Ops tf;
@@ -55,8 +55,8 @@ protected void setTF(Ops tf) {
    * Gets the calculation operation for the activation.
    *
    * @param input the input tensor
-   * @param <T> the data type of the input and result
+   * @param <U> the data type of the input and result
    * @return The operand for the activation
    */
-  public abstract <T extends TNumber> Operand<T> call(Operand<T> input);
+  public abstract <U extends T> Operand<U> call(Operand<U> input);
 }
diff --git a/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/ELU.java b/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/ELU.java
index 81b34ca77e3..00c720c936b 100644
--- a/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/ELU.java
+++ b/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/ELU.java
@@ -18,7 +18,6 @@
 import org.tensorflow.op.Ops;
 import org.tensorflow.types.TBool;
 import org.tensorflow.types.family.TFloating;
-import org.tensorflow.types.family.TNumber;
 
 /**
  * Exponential linear unit.
@@ -48,8 +47,7 @@
  * @see <a href="https://arxiv.org/abs/1511.07289">Clevert et al, 2016, Fast and Accurate Deep
  *     Network Learning by Exponential Linear Units (ELUs)</a>
  */
-// TFloating
-public class ELU extends Activation {
+public class ELU extends Activation<TFloating> {
 
   private static final double ALPHA_DEFAULT = 1.0;
 
@@ -79,19 +77,14 @@ public ELU(Ops tf, double alpha) {
 
   /** {@inheritDoc} */
   @Override
-  public <T extends TNumber> Operand<T> call(Operand<T> input) {
+  public <U extends TFloating> Operand<U> call(Operand<U> input) {
 
-    if (!TFloating.class.isAssignableFrom(input.type()) ) {
-      throw new IllegalArgumentException(
-              "Tensor type must be numeric or boolean: " + input.type().getSimpleName());
-    }
-
-    Operand<T> result = tf.nn.elu(input);
+    Operand<U> result = tf.nn.elu(input);
     if (alpha == 1.0) {
       return result;
     } else {
-      Class<T> inputType = input.type();
-      Operand<T> y = tf.math.mul(result, tf.dtypes.cast(tf.constant(alpha), inputType));
+      Class<U> inputType = input.type();
+      Operand<U> y = tf.math.mul(result, tf.dtypes.cast(tf.constant(alpha), inputType));
       Operand<TBool> cond = tf.math.greater(result, tf.dtypes.cast(tf.constant(0), inputType));
       return tf.select(cond, result, y);
     }
diff --git a/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Exponential.java b/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Exponential.java
index 6638b9f2e90..79857dd3f7b 100644
--- a/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Exponential.java
+++ b/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Exponential.java
@@ -17,7 +17,6 @@
 import org.tensorflow.Operand;
 import org.tensorflow.op.Ops;
 import org.tensorflow.types.family.TFloating;
-import org.tensorflow.types.family.TNumber;
 
 /**
  * Exponential activation function.
@@ -32,8 +31,7 @@
  *   // result is [0.04978707f,  0.36787945f,  1.f,  2.7182817f, 20.085537f]
  * </pre>
  */
-// TFloating
-public class Exponential extends Activation {
+public class Exponential extends Activation<TFloating> {
 
   /**
    * Creates an Exponential activation.
@@ -48,15 +46,12 @@ public Exponential(Ops tf) {
    * Calculates the Exponential activation.
    *
    * @param input the input tensor
-   * @param <T> the data type of the input and result
+   * @param <U> the data type of the input and result
    * @return an Operand for the exponential activation: <code>exp(x)</code>.
    */
   @Override
-  public <T extends TNumber> Operand<T> call(Operand<T> input) {
-    if (!TFloating.class.isAssignableFrom(input.type()) ) {
-      throw new IllegalArgumentException(
-              "Tensor type must be numeric or boolean: " + input.type().getSimpleName());
-    }
+  public <U extends TFloating> Operand<U> call(Operand<U> input) {
+
 
     return tf.math.exp(input);
   }
diff --git a/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/GeLU.java b/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/GeLU.java
index fa267114781..6b27dd4c12a 100644
--- a/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/GeLU.java
+++ b/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/GeLU.java
@@ -15,13 +15,8 @@
 package org.tensorflow.framework.activations;
 
 import org.tensorflow.Operand;
-import org.tensorflow.Session;
 import org.tensorflow.op.Ops;
-import org.tensorflow.types.TFloat64;
 import org.tensorflow.types.family.TFloating;
-import org.tensorflow.types.family.TNumber;
-
-import java.util.Arrays;
 
 import static org.tensorflow.framework.utils.CastHelper.cast;
 
@@ -39,7 +34,7 @@
  * <p>or, if <code>approximate</code> is <code>false</code>.
  *
  * <pre>
- *     x * P(X <= x) = 0.5 * x * (1 + erf(x / sqrt(2))),
+ *     x * P(X &lt;= x) = 0.5 * x * (1 + erf(x / sqrt(2))),
  * </pre>
  *
  * where <code>P(X) ~ N(0, 1)</code>.
@@ -47,8 +42,7 @@
  * @see <a href="https://arxiv.org/abs/1606.08415">Hendrycks, Dan and Gimpel, Kevin, 2016-2020,
  *     Gaussian Error Linear Units (GELUs)</a>
  */
-// TFloating
-public class GeLU extends Activation {
+public class GeLU extends Activation<TFloating> {
 
   private final boolean approximate;
 
@@ -74,11 +68,8 @@ public GeLU(Ops tf, boolean approximate) {
 
   /** {@inheritDoc} */
   @Override
-  public <T extends TNumber> Operand<T> call(Operand<T> input) {
-    if (!TFloating.class.isAssignableFrom(input.type())) {
-      throw new IllegalArgumentException(
-          "Tensor type must be numeric or boolean: " + input.type().getSimpleName());
-    }
+  public <U extends TFloating> Operand<U> call(Operand<U> input) {
+
     if (approximate) {
       /*
               coeff = math_ops.cast(0.044715, features.dtype)
@@ -86,9 +77,9 @@ public <T extends TNumber> Operand<T> call(Operand<T> input) {
                     1.0 + math_ops.tanh(0.7978845608028654 *
                               (features + coeff * math_ops.pow(features, 3))))
       */
-      Operand<T> coeff = cast(tf, tf.constant(0.044715), input.type());
-      Operand<T> point5 = cast(tf, tf.constant(0.5), input.type());
-      Operand<T> one = cast(tf, tf.constant(1.0), input.type());
+      Operand<U> coeff = cast(tf, tf.constant(0.044715), input.type());
+      Operand<U> point5 = cast(tf, tf.constant(0.5), input.type());
+      Operand<U> one = cast(tf, tf.constant(1.0), input.type());
 
       return tf.math.mul(
           point5,
@@ -128,5 +119,4 @@ public <T extends TNumber> Operand<T> call(Operand<T> input) {
                           input, cast(tf, tf.constant(1.4142135623730951), input.type()))))));
     }
   }
-
 }
diff --git a/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/HardSigmoid.java b/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/HardSigmoid.java
index bc8f5b55291..6c7323a90cb 100644
--- a/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/HardSigmoid.java
+++ b/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/HardSigmoid.java
@@ -41,7 +41,7 @@
  *     // result is [0.f , 0.3f, 0.5f, 0.7f, 1.f]
  * </pre>
  */
-public class HardSigmoid extends Activation {
+public class HardSigmoid extends Activation<TNumber> {
 
   /**
    * Creates Hard sigmoid activation.
@@ -54,12 +54,12 @@ public HardSigmoid(Ops tf) {
 
   /** {@inheritDoc} */
   @Override
-  public <T extends TNumber> Operand<T> call(Operand<T> input) {
-    Class<T> inputType = input.type();
-    Operand<T> point2 = tf.dtypes.cast(tf.constant(0.2), inputType);
-    Operand<T> point5 = tf.dtypes.cast(tf.constant(0.5), inputType);
+  public <U extends TNumber> Operand<U> call(Operand<U> input) {
+    Class<U> inputType = input.type();
+    Operand<U> point2 = tf.dtypes.cast(tf.constant(0.2), inputType);
+    Operand<U> point5 = tf.dtypes.cast(tf.constant(0.5), inputType);
 
-    Operand<T> x = tf.math.add(tf.math.mul(input, point2), point5);
+    Operand<U> x = tf.math.add(tf.math.mul(input, point2), point5);
     return tf.clipByValue(
         x, tf.dtypes.cast(tf.constant(0), inputType), tf.dtypes.cast(tf.constant(1), inputType));
   }
diff --git a/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Linear.java b/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Linear.java
index 06aba774423..dcda76db6bf 100644
--- a/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Linear.java
+++ b/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Linear.java
@@ -33,7 +33,7 @@
  *    // result is [-3.0f,-1.0f, 0.0f,1.0f,3.0f]
  * </pre>
  */
-public class Linear extends Activation {
+public class Linear extends Activation<TNumber> {
 
   /**
    * Creates a linear activation.
@@ -46,7 +46,7 @@ public Linear(Ops tf) {
 
   /** {@inheritDoc} */
   @Override
-  public <T extends TNumber> Operand<T> call(Operand<T> input) {
+  public <U extends TNumber> Operand<U> call(Operand<U> input) {
     return input;
   }
 }
diff --git a/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/ReLU.java b/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/ReLU.java
index b3654093109..409015a1203 100644
--- a/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/ReLU.java
+++ b/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/ReLU.java
@@ -56,8 +56,7 @@
  *     // result is [-0.f, -0.f,  0.f,  0.f, 10.f]
  * </pre>
  */
-// TFloating
-public class ReLU extends Activation {
+public class ReLU extends Activation<TNumber> {
 
   public static final float ALPHA_DEFAULT = 0.0f;
   public static final float MAX_VALUE_DEFAULT = Float.NaN;
@@ -95,11 +94,11 @@ public ReLU(Ops tf, float alpha, float maxValue, float threshold) {
 
   /** {@inheritDoc} */
   @Override
-  public <T extends TNumber> Operand<T> call(Operand<T> input) {
-    Class<T> inputType = input.type();
+  public <U extends TNumber> Operand<U> call(Operand<U> input) {
+    Class<U> inputType = input.type();
 
     boolean clipMax = !Float.isNaN(maxValue);
-    Operand<T> negativePart = null;
+    Operand<U> negativePart = null;
     if (alpha != 0) {
       if (Float.isNaN(maxValue) && threshold == 0) {
         return tf.nn.leakyRelu(input, LeakyRelu.alpha(alpha));
@@ -113,7 +112,7 @@ public <T extends TNumber> Operand<T> call(Operand<T> input) {
       }
     }
 
-    Operand<T> lInput;
+    Operand<U> lInput;
     if (threshold != 0) {
       // computes input for input > threshold else 0
       Greater greater = tf.math.greater(input, tf.dtypes.cast(tf.constant(threshold), inputType));
@@ -126,8 +125,8 @@ public <T extends TNumber> Operand<T> call(Operand<T> input) {
       lInput = tf.nn.relu(input);
     }
     if (clipMax) {
-      Operand<T> lmaxValue = tf.dtypes.cast(tf.constant(maxValue), inputType);
-      Operand<T> zero = tf.dtypes.cast(tf.constant(0), inputType);
+      Operand<U> lmaxValue = tf.dtypes.cast(tf.constant(maxValue), inputType);
+      Operand<U> zero = tf.dtypes.cast(tf.constant(0), inputType);
       lInput = tf.clipByValue(lInput, zero, lmaxValue);
     }
 
diff --git a/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/SELU.java b/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/SELU.java
index 0d94d3c5092..7955144abbb 100644
--- a/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/SELU.java
+++ b/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/SELU.java
@@ -16,7 +16,7 @@
 
 import org.tensorflow.Operand;
 import org.tensorflow.op.Ops;
-import org.tensorflow.types.family.TNumber;
+import org.tensorflow.types.family.TFloating;
 
 /**
  * Scaled Exponential Linear Unit (SELU).
@@ -44,8 +44,7 @@
  *
  * @see <a href="https://arxiv.org/abs/1706.02515">Klambauer et al., 2017</a>
  */
-// TFloating
-public class SELU extends Activation {
+public class SELU extends Activation<TFloating> {
 
   /**
    * Creates a Scaled Exponential Linear Unit (SELU) activation.
@@ -58,7 +57,7 @@ public SELU(Ops tf) {
 
   /** {@inheritDoc} */
   @Override
-  public <T extends TNumber> Operand<T> call(Operand<T> input) {
+  public <U extends TFloating> Operand<U> call(Operand<U> input) {
     return tf.nn.selu(input);
   }
 }
diff --git a/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Sigmoid.java b/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Sigmoid.java
index eca6d2c3f8c..a89b6119d02 100644
--- a/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Sigmoid.java
+++ b/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Sigmoid.java
@@ -16,7 +16,7 @@
 
 import org.tensorflow.Operand;
 import org.tensorflow.op.Ops;
-import org.tensorflow.types.family.TNumber;
+import org.tensorflow.types.family.TFloating;
 
 /**
  * Sigmoid activation. <code>sigmoid(x) = 1 / (1 + exp(-x))</code>.
@@ -39,8 +39,7 @@
  *     //                 5.0000000e-01f,7.3105860e-01f, 1.f]
  * </pre>
  */
-// TFloating
-public class Sigmoid extends Activation {
+public class Sigmoid extends Activation<TFloating> {
 
   /**
    * Creates a Sigmoid activation.
@@ -53,7 +52,7 @@ public Sigmoid(Ops tf) {
 
   /** {@inheritDoc} */
   @Override
-  public <T extends TNumber> Operand<T> call(Operand<T> input) {
+  public <U extends TFloating> Operand<U> call(Operand<U> input) {
     return tf.math.sigmoid(input);
   }
 }
diff --git a/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Softmax.java b/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Softmax.java
index 6991d7d233c..309cdd68b8b 100644
--- a/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Softmax.java
+++ b/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Softmax.java
@@ -19,7 +19,7 @@
 import org.tensorflow.op.Ops;
 import org.tensorflow.op.core.ReduceMax;
 import org.tensorflow.op.core.ReduceSum;
-import org.tensorflow.types.family.TNumber;
+import org.tensorflow.types.family.TFloating;
 
 /**
  * Softmax converts a real vector to a vector of categorical probabilities.
@@ -36,8 +36,7 @@
  *
  * <p>The input values in are the log-odds of the resulting probability.
  */
-// TFloating
-public class Softmax extends Activation {
+public class Softmax extends Activation<TFloating> {
 
   private static final int AXIS_DEFAULT = -1;
 
@@ -66,16 +65,16 @@ public Softmax(Ops tf, int axis) {
 
   /** {@inheritDoc} */
   @Override
-  public <T extends TNumber> Operand<T> call(Operand<T> input) {
+  public <U extends TFloating> Operand<U> call(Operand<U> input) {
     Shape shape = input.shape();
     int numDimensions = shape.numDimensions();
     if (numDimensions == 2) {
       return tf.nn.softmax(input);
     } else {
-      Operand<T> e =
+      Operand<U> e =
           tf.math.exp(
               tf.math.sub(input, tf.reduceMax(input, tf.constant(axis), ReduceMax.keepDims(true))));
-      Operand<T> s = tf.reduceSum(e, tf.constant(axis), ReduceSum.keepDims(true));
+      Operand<U> s = tf.reduceSum(e, tf.constant(axis), ReduceSum.keepDims(true));
       return tf.math.div(e, s);
     }
   }
diff --git a/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Softplus.java b/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Softplus.java
index f34d756afda..0eb703aad9f 100644
--- a/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Softplus.java
+++ b/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Softplus.java
@@ -16,7 +16,7 @@
 
 import org.tensorflow.Operand;
 import org.tensorflow.op.Ops;
-import org.tensorflow.types.family.TNumber;
+import org.tensorflow.types.family.TFloating;
 
 /**
  * Softplus activation function, <code>softplus(x) = log(exp(x) + 1)</code>.
@@ -32,8 +32,7 @@
  *     //                 1.3132616e+00f, 2.0000000e+01f]
  * </pre>
  */
-// TFloating
-public class Softplus extends Activation {
+public class Softplus extends Activation<TFloating> {
 
   /**
    * Creates a Softplus activation function.
@@ -46,7 +45,7 @@ public Softplus(Ops tf) {
 
   /** {@inheritDoc} */
   @Override
-  public <T extends TNumber> Operand<T> call(Operand<T> input) {
+  public <U extends TFloating> Operand<U> call(Operand<U> input) {
     return tf.math.softplus(input);
   }
 }
diff --git a/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Softsign.java b/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Softsign.java
index db308154e2f..0a7754258df 100644
--- a/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Softsign.java
+++ b/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Softsign.java
@@ -16,7 +16,7 @@
 
 import org.tensorflow.Operand;
 import org.tensorflow.op.Ops;
-import org.tensorflow.types.family.TNumber;
+import org.tensorflow.types.family.TFloating;
 
 /**
  * Softsign activation function, <code>softsign(x) = x / (abs(x) + 1)</code>.
@@ -31,8 +31,7 @@
  *     // result is [-0.5f, 0.f, 0.5f]
  * </pre>
  */
-// TFloating
-public class Softsign extends Activation {
+public class Softsign extends Activation<TFloating> {
 
   /**
    * Creates a Softsign activation.
@@ -45,7 +44,7 @@ public Softsign(Ops tf) {
 
   /** {@inheritDoc} */
   @Override
-  public <T extends TNumber> Operand<T> call(Operand<T> input) {
+  public <U extends TFloating> Operand<U> call(Operand<U> input) {
     return tf.nn.softsign(input);
   }
 }
diff --git a/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Swish.java b/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Swish.java
index 233e9016b56..eb43a21f285 100644
--- a/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Swish.java
+++ b/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Swish.java
@@ -16,7 +16,7 @@
 
 import org.tensorflow.Operand;
 import org.tensorflow.op.Ops;
-import org.tensorflow.types.family.TNumber;
+import org.tensorflow.types.family.TFloating;
 
 /**
  * Swish activation function. <code>swish(x) = x * sigmoid(x)</code>.
@@ -39,8 +39,7 @@
  *
  * @see <a href="https://arxiv.org/abs/1710.05941">Ramachandran et al., 2017</a>
  */
-// TFloating
-public class Swish extends Activation {
+public class Swish extends Activation<TFloating> {
 
   /**
    * Creates a Swish activation, <code>swish(x) = x * sigmoid(x)</code>.
@@ -57,7 +56,7 @@ public Swish(Ops tf) {
 
   /** {@inheritDoc} */
   @Override
-  public <T extends TNumber> Operand<T> call(Operand<T> input) {
+  public <U extends TFloating> Operand<U> call(Operand<U> input) {
 
     // TODO Python Keras returns a "grad", which is an optimization not implemented in Java.
     return tf.math.mul(input, tf.math.sigmoid(input));
diff --git a/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Tanh.java b/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Tanh.java
index b4d3499ca6f..145561b9129 100644
--- a/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Tanh.java
+++ b/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Tanh.java
@@ -16,7 +16,7 @@
 
 import org.tensorflow.Operand;
 import org.tensorflow.op.Ops;
-import org.tensorflow.types.family.TNumber;
+import org.tensorflow.types.family.TFloating;
 
 /**
  * Hyperbolic tangent activation function.
@@ -31,8 +31,7 @@
  *     // result = [-0.9950547f, -0.7615942f,  0.f,  0.7615942f,  0.9950547f]
  * </pre>
  */
-// TFloating
-public class Tanh extends Activation {
+public class Tanh extends Activation<TFloating> {
 
   /**
    * Creates a Hyperbolic tangent activation.
@@ -45,7 +44,7 @@ public Tanh(Ops tf) {
 
   /** {@inheritDoc} */
   @Override
-  public <T extends TNumber> Operand<T> call(Operand<T> input) {
+  public <U extends TFloating> Operand<U> call(Operand<U> input) {
     return tf.math.tanh(input);
   }
 }
diff --git a/tensorflow-framework/src/main/java/org/tensorflow/framework/constraints/Constraint.java b/tensorflow-framework/src/main/java/org/tensorflow/framework/constraints/Constraint.java
index d3094b5e9e9..739f9f55c55 100644
--- a/tensorflow-framework/src/main/java/org/tensorflow/framework/constraints/Constraint.java
+++ b/tensorflow-framework/src/main/java/org/tensorflow/framework/constraints/Constraint.java
@@ -41,7 +41,9 @@ public Constraint(Ops tf) {
    * Applies the constraint against the provided weights
    *
    * @param weights the weights
+   * @param <T> the data the weights and result
    * @return the constrained weights
+   *
    */
   public abstract <T extends TNumber> Operand<T> call(Operand<T> weights);
 
diff --git a/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/BaseInitializer.java b/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/BaseInitializer.java
index 7efd02a6db1..9c1fa9ac287 100644
--- a/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/BaseInitializer.java
+++ b/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/BaseInitializer.java
@@ -15,9 +15,10 @@
 package org.tensorflow.framework.initializers;
 
 import org.tensorflow.op.Ops;
+import org.tensorflow.types.family.TType;
 
 /** Abstract base class for all Initializers */
-public abstract class BaseInitializer implements Initializer {
+public abstract class BaseInitializer<T extends TType> implements Initializer<T> {
 
   protected final Ops tf;
 
diff --git a/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Identity.java b/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Identity.java
index 560aa394454..2b1ccf00bae 100644
--- a/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Identity.java
+++ b/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Identity.java
@@ -19,7 +19,7 @@
 import org.tensorflow.ndarray.Shape;
 import org.tensorflow.op.Ops;
 import org.tensorflow.types.TInt64;
-import org.tensorflow.types.family.TType;
+import org.tensorflow.types.family.TFloating;
 
 import static org.tensorflow.framework.utils.CastHelper.cast;
 
@@ -37,7 +37,7 @@
  *             initializer.call(tf.constant(Shape.of(2,2)), TFloat32.class);
  * </pre>
  */
-public class Identity extends BaseInitializer {
+public class Identity extends BaseInitializer<TFloating> {
   public static final double GAIN_DEFAULT = 1.0;
 
   private final double gain;
@@ -65,7 +65,7 @@ public Identity(Ops tf, double gain) {
 
   /** {@inheritDoc} */
   @Override
-  public <T extends TType> Operand<T> call(Operand<TInt64> dims, Class<T> type) {
+  public <U extends TFloating> Operand<U> call(Operand<TInt64> dims, Class<U> type) {
     Shape shape = ShapeUtils.toShape(tf.scope(), dims);
     if (shape.numDimensions() != 2) {
       throw new IllegalArgumentException("2D matrix required, got " + shape.numDimensions());
@@ -74,9 +74,9 @@ public <T extends TType> Operand<T> call(Operand<TInt64> dims, Class<T> type) {
     long diagSize = Math.min(shape.size(0), shape.size(1));
     Shape diagShape = Shape.of(diagSize);
 
-    Operand<T> op;
-    Operand<T> zero = cast(tf, tf.constant(0), type);
-    Operand<T> diagOnes =
+    Operand<U> op;
+    Operand<U> zero = cast(tf, tf.constant(0), type);
+    Operand<U> diagOnes =
         tf.fill(tf.constant(diagShape.asArray()), cast(tf, tf.constant(1.0), type));
     if (isSquare) {
       op =
@@ -87,7 +87,7 @@ public <T extends TType> Operand<T> call(Operand<TInt64> dims, Class<T> type) {
               tf.constant((int) shape.size(1)),
               zero);
     } else {
-      Operand<T> zeroMatrix = tf.zeros(dims, type);
+      Operand<U> zeroMatrix = tf.zeros(dims, type);
       op = tf.linalg.matrixSetDiag(zeroMatrix, diagOnes, tf.constant(0));
     }
 
diff --git a/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Initializer.java b/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Initializer.java
index c22a2582d35..2f30bc5e99d 100644
--- a/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Initializer.java
+++ b/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Initializer.java
@@ -19,15 +19,15 @@
 import org.tensorflow.types.family.TType;
 
 /** An interface for Initializers */
-public interface Initializer {
+public interface Initializer<T extends TType> {
 
   /**
    * Generates the operation used to perform the initialization.
    *
    * @param dims the shape dimensions
    * @param type the data type of tensor
-   * @param <T> The data Type for initializer operation
+   * @param <U> The data Type for initializer operation
    * @return An operand for the initialization.
    */
-  <T extends TType> Operand<T> call(Operand<TInt64> dims, Class<T> type);
+  <U extends T> Operand<U> call(Operand<TInt64> dims, Class<U> type);
 }
diff --git a/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Ones.java b/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Ones.java
index 52528df784f..e2ad49c07f9 100644
--- a/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Ones.java
+++ b/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Ones.java
@@ -35,7 +35,7 @@
  *              initializer.call(tf.constant(Shape.of(2,2)), TFloat32.class);
  * </pre>
  */
-public class Ones extends BaseInitializer {
+public class Ones extends BaseInitializer<TType> {
 
   /**
    * Creates an Initializer that sets all values to one.
@@ -55,9 +55,17 @@ public Ones(Ops tf) {
     super(tf);
   }
 
-  /** {@inheritDoc} */
+  /**
+   * Generates the operation used to perform the initialization.
+   *
+   * @param dims the shape dimensions
+   * @param type the data type of tensor
+   * @param <U> The data Type for initializer operation
+   * @return An operand for the initialization.
+   * @throws IllegalArgumentException if the data type is not a TNumber or TBool
+   */
   @Override
-  public <T extends TType> Operand<T> call(Operand<TInt64> dims, Class<T> type) {
+  public <U extends TType> Operand<U> call(Operand<TInt64> dims, Class<U> type) {
     if (!TNumber.class.isAssignableFrom(type) && type != TBool.class) {
       throw new IllegalArgumentException(
           "Tensor type must be numeric or boolean: " + type.getSimpleName());
diff --git a/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Orthogonal.java b/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Orthogonal.java
index de5fb825e42..14f1049d038 100644
--- a/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Orthogonal.java
+++ b/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Orthogonal.java
@@ -21,8 +21,7 @@
 import org.tensorflow.op.Ops;
 import org.tensorflow.op.linalg.Qr;
 import org.tensorflow.types.TInt64;
-import org.tensorflow.types.family.TNumber;
-import org.tensorflow.types.family.TType;
+import org.tensorflow.types.family.TFloating;
 
 import static org.tensorflow.framework.utils.CastHelper.cast;
 
@@ -48,7 +47,7 @@
  *              initializer.call(tf.constant(Shape.of(2,2)), TFloat32.class);
  * </pre>
  */
-public class Orthogonal extends BaseInitializer {
+public class Orthogonal extends BaseInitializer<TFloating> {
 
   public static final double GAIN_DEFAULT = 1.0;
 
@@ -82,11 +81,8 @@ public Orthogonal(Ops tf, double gain, long seed) {
 
   /** {@inheritDoc} */
   @Override
-  public <T extends TType> Operand<T> call(Operand<TInt64> dims, Class<T> type) {
-    if (!TNumber.class.isAssignableFrom(type)) {
-      throw new IllegalArgumentException("Tensor type must be numeric: " + type.getSimpleName());
-    }
-    Class<TNumber> nType = (Class<TNumber>) type;
+  public <U extends TFloating> Operand<U> call(Operand<TInt64> dims, Class<U> type) {
+
     Shape dimsShape = ShapeUtils.toShape(tf.scope(), dims);
     if (dimsShape.numDimensions() < 2) {
       throw new IllegalArgumentException(
@@ -99,18 +95,20 @@ public <T extends TType> Operand<T> call(Operand<TInt64> dims, Class<T> type) {
     long numCols = dimsShape.size(i);
     Shape flatShape = Shape.of(Math.max(numRows, numCols), Math.min(numRows, numCols));
     long[] seeds = {seed, 0};
-    Operand<TNumber> op =
-        tf.random.statelessRandomNormal(tf.constant(flatShape), tf.constant(seeds), nType);
+    Operand<U> op =
+        tf.random.statelessRandomNormal(tf.constant(flatShape), tf.constant(seeds), type);
 
     Qr.Options qrOptions = Qr.fullMatrices(false);
-    Qr<TNumber> qrOp = tf.linalg.qr(op, qrOptions);
-    Output<TNumber> qo = qrOp.q();
-    Output<TNumber> ro = qrOp.r();
-    Operand<TNumber> diagOp =
+    Qr<U> qrOp = tf.linalg.qr(op, qrOptions);
+    Output<U> qo = qrOp.q();
+    Output<U> ro = qrOp.r();
+    Operand<U> diagOp =
         tf.linalg.matrixDiagPart(ro, tf.constant(0), cast(tf, tf.constant(0), op.type()));
-    Operand<TNumber> qop = tf.math.mul(qo, tf.math.sign(diagOp));
-    if (numRows < numCols) qop = tf.linalg.transpose(qop, null);
+    Operand<U> qop = tf.math.mul(qo, tf.math.sign(diagOp));
+    if (numRows < numCols) {
+      qop = tf.linalg.transpose(qop, null);
+    }
 
-    return cast(tf, tf.math.mul(qop, cast(tf, tf.constant(this.gain), op.type())), type);
+    return tf.math.mul(qop, cast(tf, tf.constant(this.gain), op.type()));
   }
 }
diff --git a/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/RandomNormal.java b/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/RandomNormal.java
index 7c5d2f9a17b..789942f52ce 100644
--- a/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/RandomNormal.java
+++ b/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/RandomNormal.java
@@ -17,8 +17,7 @@
 import org.tensorflow.Operand;
 import org.tensorflow.op.Ops;
 import org.tensorflow.types.TInt64;
-import org.tensorflow.types.family.TNumber;
-import org.tensorflow.types.family.TType;
+import org.tensorflow.types.family.TFloating;
 
 import static org.tensorflow.framework.utils.CastHelper.cast;
 
@@ -35,7 +34,7 @@
  *              initializer.call(tf.constant(Shape.of(2,2)), TFloat32.class);
  * </pre>
  */
-public class RandomNormal extends BaseInitializer {
+public class RandomNormal extends BaseInitializer<TFloating> {
 
   public static final double MEAN_DEFAULT = 0.0;
   public static final double STDDEV_DEFAULT = 1.0;
@@ -86,17 +85,10 @@ public RandomNormal(Ops tf, double mean, double stddev, long seed) {
 
   /** {@inheritDoc} */
   @Override
-  public <T extends TType> Operand<T> call(Operand<TInt64> dims, Class<T> type) {
-    if (!TNumber.class.isAssignableFrom(type)) {
-      throw new IllegalArgumentException("Tensor type must be numeric: " + type.getSimpleName());
-    }
+  public <U extends TFloating> Operand<U> call(Operand<TInt64> dims, Class<U> type) {
     long[] seeds = {seed, 0};
-    // Suppression is ok because it is guarded by the if statement
-    @SuppressWarnings("unchecked")
-    Class<TNumber> nType = (Class<TNumber>) type;
-    Operand<TNumber> distOp = tf.random.statelessRandomNormal(dims, tf.constant(seeds), nType);
-    Operand<TNumber> op =
-        tf.math.mul(distOp, cast(tf, tf.constant(this.stddev), distOp.type()));
-    return cast(tf, tf.math.add(op, cast(tf, tf.constant(mean), distOp.type())), type);
+    Operand<U> distOp = tf.random.statelessRandomNormal(dims, tf.constant(seeds), type);
+    Operand<U> op = tf.math.mul(distOp, cast(tf, tf.constant(this.stddev), type));
+    return tf.math.add(op, cast(tf, tf.constant(mean), distOp.type()));
   }
 }
diff --git a/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/RandomUniform.java b/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/RandomUniform.java
index 30bbd61bae7..c0ebce135e9 100644
--- a/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/RandomUniform.java
+++ b/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/RandomUniform.java
@@ -20,7 +20,6 @@
 import org.tensorflow.types.TInt64;
 import org.tensorflow.types.family.TIntegral;
 import org.tensorflow.types.family.TNumber;
-import org.tensorflow.types.family.TType;
 
 import static org.tensorflow.framework.utils.CastHelper.cast;
 
@@ -37,7 +36,7 @@
  *              initializer.call(tf.constant(Shape.of(2,2)), TFloat32.class);
  * </pre>
  */
-public class RandomUniform extends BaseInitializer {
+public class RandomUniform extends BaseInitializer<TNumber> {
 
   public static final double MINVAL_DEFAULT = -0.05;
   public static final double MAXVAL_DEFAULT = 0.05;
@@ -76,25 +75,19 @@ public RandomUniform(Ops tf, double minval, double maxval, long seed) {
 
   /** {@inheritDoc} */
   @Override
-  public <T extends TType> Operand<T> call(Operand<TInt64> dims, Class<T> type) {
-    if (!TNumber.class.isAssignableFrom(type)) {
-      throw new IllegalArgumentException("Tensor type must be numeric: " + type.getSimpleName());
-    }
-    // Suppression is ok because it is guarded by the if statement
-    @SuppressWarnings("unchecked")
-    Class<TNumber> nType = (Class<TNumber>) type;
-    Operand<TNumber> distOp;
+  public <U extends TNumber> Operand<U> call(Operand<TInt64> dims, Class<U> type) {
+    Operand<U> distOp;
     if (TIntegral.class.isAssignableFrom(type)) {
       RandomUniformInt.Options options = RandomUniformInt.seed(this.seed);
       distOp =
           tf.random.randomUniformInt(
               dims,
-              cast(tf, tf.constant(this.minval), nType),
-              cast(tf, tf.constant(this.maxval), nType),
+              cast(tf, tf.constant(this.minval), type),
+              cast(tf, tf.constant(this.maxval), type),
               options);
     } else {
       long[] seeds = {seed, 0};
-      distOp = tf.random.statelessRandomUniform(dims, tf.constant(seeds), nType);
+      distOp = tf.random.statelessRandomUniform(dims, tf.constant(seeds), type);
       if (this.minval == 0) {
         if (this.maxval != 1.0) {
           distOp = tf.math.mul(distOp, cast(tf, tf.constant(this.maxval), distOp.type()));
@@ -105,6 +98,6 @@ public <T extends TType> Operand<T> call(Operand<TInt64> dims, Class<T> type) {
         distOp = tf.math.add(distOp, cast(tf, tf.constant(this.minval), distOp.type()));
       }
     }
-    return cast(tf, distOp, type);
+    return distOp;
   }
 }
diff --git a/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/TruncatedNormal.java b/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/TruncatedNormal.java
index 2e53c2db6f9..eaf94663993 100644
--- a/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/TruncatedNormal.java
+++ b/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/TruncatedNormal.java
@@ -17,8 +17,7 @@
 import org.tensorflow.Operand;
 import org.tensorflow.op.Ops;
 import org.tensorflow.types.TInt64;
-import org.tensorflow.types.family.TNumber;
-import org.tensorflow.types.family.TType;
+import org.tensorflow.types.family.TFloating;
 
 import static org.tensorflow.framework.utils.CastHelper.cast;
 
@@ -35,7 +34,7 @@
  *              initializer.call(tf.constant(Shape.of(2,2)), TFloat32.class);
  * </pre>
  */
-public class TruncatedNormal extends BaseInitializer {
+public class TruncatedNormal extends BaseInitializer<TFloating> {
 
   public static final double MEAN_DEFAULT = 0.0;
   public static final double STDDEV_DEFAULT = 0.05;
@@ -74,20 +73,11 @@ public TruncatedNormal(Ops tf, double mean, double stddev, long seed) {
 
   /** {@inheritDoc} */
   @Override
-  public <T extends TType> Operand<T> call(Operand<TInt64> dims, Class<T> type) {
-    if (!TNumber.class.isAssignableFrom(type)) {
-      throw new IllegalArgumentException("Tensor type must be numeric: " + type.getSimpleName());
-    }
-    // Suppression is ok because it is guarded by the if statement
-    @SuppressWarnings("unchecked")
-    Class<TNumber> nType = (Class<TNumber>) type;
+  public <U extends TFloating> Operand<U> call(Operand<TInt64> dims, Class<U> type) {
     long[] seeds = {seed, 0};
-    Operand<TNumber> distOp = tf.random.statelessTruncatedNormal(dims, tf.constant(seeds), nType);
-    return cast(
-        tf,
-        tf.math.add(
-            tf.math.mul(distOp, cast(tf, tf.constant(stddev), distOp.type())),
-            cast(tf, tf.constant(mean), distOp.type())),
-        type);
+    Operand<U> distOp = tf.random.statelessTruncatedNormal(dims, tf.constant(seeds), type);
+    return tf.math.add(
+        tf.math.mul(distOp, cast(tf, tf.constant(stddev), distOp.type())),
+        cast(tf, tf.constant(mean), distOp.type()));
   }
 }
diff --git a/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/VarianceScaling.java b/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/VarianceScaling.java
index d820f04a170..299d719c7c3 100644
--- a/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/VarianceScaling.java
+++ b/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/VarianceScaling.java
@@ -19,8 +19,7 @@
 import org.tensorflow.ndarray.Shape;
 import org.tensorflow.op.Ops;
 import org.tensorflow.types.TInt64;
-import org.tensorflow.types.family.TNumber;
-import org.tensorflow.types.family.TType;
+import org.tensorflow.types.family.TFloating;
 
 import static org.tensorflow.framework.utils.CastHelper.cast;
 
@@ -55,7 +54,7 @@
  * @see VarianceScaling.Mode
  * @see VarianceScaling.Distribution
  */
-public class VarianceScaling extends BaseInitializer {
+public class VarianceScaling extends BaseInitializer<TFloating> {
 
   public static final double SCALE_DEFAULT = 1.0;
   public static final Mode MODE_DEFAULT = Mode.FAN_IN;
@@ -98,13 +97,7 @@ public VarianceScaling(Ops tf, double scale, Mode mode, Distribution distributio
 
   /** {@inheritDoc} */
   @Override
-  public <T extends TType> Operand<T> call(Operand<TInt64> dims, Class<T> type) {
-    if (!TNumber.class.isAssignableFrom(type)) {
-      throw new IllegalArgumentException("Tensor type must be numeric: " + type.getSimpleName());
-    }
-    // Suppression is ok because it is guarded by the if statement
-    @SuppressWarnings("unchecked")
-    Class<TNumber> nType = (Class<TNumber>) type;
+  public <U extends TFloating> Operand<U> call(Operand<TInt64> dims, Class<U> type) {
     Shape shape = ShapeUtils.toShape(this.tf.scope(), dims);
     double lscale = this.scale;
     double[] fans /* fanIn, fanOut */ = computeFans(shape);
@@ -119,28 +112,28 @@ public <T extends TType> Operand<T> call(Operand<TInt64> dims, Class<T> type) {
         lscale /= Math.max(1., (fans[0] + fans[1]) / 2.);
         break;
     }
-    Operand<TNumber> distOp;
-    Operand<TNumber> mulOp = null;
+    Operand<U> distOp;
+    Operand<U> mulOp = null;
     double stddev;
     long[] seeds = {seed, 0};
     switch (distribution) {
       case TRUNCATED_NORMAL:
-        distOp = tf.random.statelessTruncatedNormal(dims, tf.constant(seeds), nType);
+        distOp = tf.random.statelessTruncatedNormal(dims, tf.constant(seeds), type);
         stddev = Math.sqrt(lscale) / .87962566103423978;
-        mulOp = tf.math.mul(distOp, cast(tf, tf.constant(stddev), nType));
+        mulOp = tf.math.mul(distOp, cast(tf, tf.constant(stddev), type));
         break;
       case NORMAL:
-        distOp = tf.random.statelessRandomNormal(dims, tf.constant(seeds), nType);
+        distOp = tf.random.statelessRandomNormal(dims, tf.constant(seeds), type);
         stddev = Math.sqrt(lscale);
-        mulOp = tf.math.mul(distOp, cast(tf, tf.constant(stddev), nType));
+        mulOp = tf.math.mul(distOp, cast(tf, tf.constant(stddev), type));
         break;
       case UNIFORM:
-        distOp = tf.random.statelessRandomUniform(dims, tf.constant(seeds), nType);
+        distOp = tf.random.statelessRandomUniform(dims, tf.constant(seeds), type);
         stddev = Math.sqrt(3.0 * lscale);
-        mulOp = tf.math.mul(distOp, cast(tf, tf.constant(stddev), nType));
+        mulOp = tf.math.mul(distOp, cast(tf, tf.constant(stddev), type));
         break;
     }
-    return cast(tf, mulOp, type);
+    return mulOp;
   }
 
   /**
diff --git a/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Zeros.java b/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Zeros.java
index 8368baad804..b6487dc10cd 100644
--- a/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Zeros.java
+++ b/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Zeros.java
@@ -16,7 +16,9 @@
 
 import org.tensorflow.Operand;
 import org.tensorflow.op.Ops;
+import org.tensorflow.types.TBool;
 import org.tensorflow.types.TInt64;
+import org.tensorflow.types.family.TNumber;
 import org.tensorflow.types.family.TType;
 
 /**
@@ -31,7 +33,7 @@
  *              initializer.call(tf.constant(Shape.of(2,2)), TFloat32.class);
  * </pre>
  */
-public class Zeros extends BaseInitializer {
+public class Zeros extends BaseInitializer<TType> {
 
   /**
    * Creates an Initializer that sets all values to one.
@@ -42,9 +44,21 @@ public Zeros(Ops tf) {
     super(tf);
   }
 
-  /** {@inheritDoc} */
+  /**
+   * Generates the operation used to perform the initialization.
+   *
+   * @param dims the shape dimensions
+   * @param type the data type of tensor
+   * @param <U> The data Type for initializer operation
+   * @return An operand for the initialization.
+   * @throws IllegalArgumentException if the data type is not a TNumber or TBool
+   */
   @Override
-  public <T extends TType> Operand<T> call(Operand<TInt64> dims, Class<T> type) {
+  public <U extends TType> Operand<U> call(Operand<TInt64> dims, Class<U> type) {
+    if (!TNumber.class.isAssignableFrom(type) && type != TBool.class) {
+      throw new IllegalArgumentException(
+          "Tensor type must be numeric or boolean: " + type.getSimpleName());
+    }
     return tf.zeros(dims, type);
   }
 }
diff --git a/tensorflow-framework/src/test/java/org/tensorflow/framework/activations/ExponentialTest.java b/tensorflow-framework/src/test/java/org/tensorflow/framework/activations/ExponentialTest.java
index beb5a1d9c9d..e8758148c21 100644
--- a/tensorflow-framework/src/test/java/org/tensorflow/framework/activations/ExponentialTest.java
+++ b/tensorflow-framework/src/test/java/org/tensorflow/framework/activations/ExponentialTest.java
@@ -20,7 +20,6 @@
 import org.tensorflow.op.Ops;
 import org.tensorflow.types.TFloat32;
 import org.tensorflow.types.TFloat64;
-import org.tensorflow.types.TInt32;
 
 /** @author Jim Clarke */
 public class ExponentialTest {
@@ -40,20 +39,6 @@ public void setUp() {}
   @AfterEach
   public void tearDown() {}
 
-  /** Test of Exponential call method. */
-  @Test
-  public void testCallInt32() {
-    int[] input = {1, -2, 3, -4, -1, 2, -3, 4};
-    for (TestSession.Mode tfMode : tfModes)
-      try (TestSession session = TestSession.createTestSession(tfMode)) {
-        Ops tf = session.getTF();
-        Exponential instance = new Exponential(tf);
-        Operand<TInt32> result = instance.call(tf.constant(input));
-        session.print(result);
-        //session.evaluate(expected, result);
-      }
-  }
-
   /** Test of Exponential call method. */
   @Test
   public void testCallFloat() {
diff --git a/tensorflow-framework/src/test/java/org/tensorflow/framework/activations/GeLUTest.java b/tensorflow-framework/src/test/java/org/tensorflow/framework/activations/GeLUTest.java
index bea6549dc3e..e2c97e18bbd 100644
--- a/tensorflow-framework/src/test/java/org/tensorflow/framework/activations/GeLUTest.java
+++ b/tensorflow-framework/src/test/java/org/tensorflow/framework/activations/GeLUTest.java
@@ -7,82 +7,87 @@
 import org.tensorflow.types.TFloat32;
 import org.tensorflow.types.TFloat64;
 
-import static org.junit.jupiter.api.Assertions.*;
-
 class GeLUTest {
 
-    private final TestSession.Mode[] tfModes = {TestSession.Mode.EAGER, TestSession.Mode.GRAPH};
+  private final TestSession.Mode[] tfModes = {TestSession.Mode.EAGER, TestSession.Mode.GRAPH};
 
-    /** Test of GeLU call method */
-    @Test
-    public void testCallFloat() {
-        float[][] input = {
-        {0.22805803f, 0.60407318f, 0.91519962f, 0.35643331f, 0.28702669f},
-        {0.11558246f, 0.57658853f, 0.47569648f, 0.02271072f, 0.24709974f}};
-        float[][] expected = {{0.13459972f, 0.43922312f, 0.75042395f, 0.22784713f, 0.17593417f},
-       {0.06310898f, 0.41392788f, 0.32483157f, 0.01156111f, 0.14766297f}};
-        for (TestSession.Mode tfMode : tfModes)
-            try (TestSession session = TestSession.createTestSession(tfMode)) {
-                Ops tf = session.getTF();
-                GeLU instance = new GeLU(tf);
-                Operand<TFloat32> result = instance.call(tf.constant(input));
-                session.evaluate(tf.constant(expected), result);
-            }
-    }
+  /** Test of GeLU call method */
+  @Test
+  public void testCallFloat() {
+    float[][] input = {
+      {0.22805803f, 0.60407318f, 0.91519962f, 0.35643331f, 0.28702669f},
+      {0.11558246f, 0.57658853f, 0.47569648f, 0.02271072f, 0.24709974f}
+    };
+    float[][] expected = {
+      {0.13459972f, 0.43922312f, 0.75042395f, 0.22784713f, 0.17593417f},
+      {0.06310898f, 0.41392788f, 0.32483157f, 0.01156111f, 0.14766297f}
+    };
+    for (TestSession.Mode tfMode : tfModes)
+      try (TestSession session = TestSession.createTestSession(tfMode)) {
+        Ops tf = session.getTF();
+        GeLU instance = new GeLU(tf);
+        Operand<TFloat32> result = instance.call(tf.constant(input));
+        session.evaluate(tf.constant(expected), result);
+      }
+  }
 
-    /** Test of GeLU call method */
-    @Test
-    public void testCallDouble() {
-        double[][] input = {
-                {0.22805803, 0.60407318, 0.91519962, 0.35643331, 0.28702669},
-                {0.11558246, 0.57658853, 0.47569648, 0.02271072, 0.24709974}};
-        double[][] expected = {
-                {0.13459972, 0.43922312, 0.75042395, 0.22784713, 0.17593417},
-                {0.06310898, 0.41392788, 0.32483157, 0.01156111, 0.14766297}
-        };
-        for (TestSession.Mode tfMode : tfModes)
-            try (TestSession session = TestSession.createTestSession(tfMode)) {
-                Ops tf = session.getTF();
-                GeLU instance = new GeLU(tf);
-                Operand<TFloat64> result = instance.call(tf.constant(input));
-                session.evaluate(tf.constant(expected), result);
-            }
-    }
+  /** Test of GeLU call method */
+  @Test
+  public void testCallDouble() {
+    double[][] input = {
+      {0.22805803, 0.60407318, 0.91519962, 0.35643331, 0.28702669},
+      {0.11558246, 0.57658853, 0.47569648, 0.02271072, 0.24709974}
+    };
+    double[][] expected = {
+      {0.13459972, 0.43922312, 0.75042395, 0.22784713, 0.17593417},
+      {0.06310898, 0.41392788, 0.32483157, 0.01156111, 0.14766297}
+    };
+    for (TestSession.Mode tfMode : tfModes)
+      try (TestSession session = TestSession.createTestSession(tfMode)) {
+        Ops tf = session.getTF();
+        GeLU instance = new GeLU(tf);
+        Operand<TFloat64> result = instance.call(tf.constant(input));
+        session.evaluate(tf.constant(expected), result);
+      }
+  }
 
-    /** Test of GeLU call method */
-    @Test
-    public void testCallFloatApproximate() {
-        float[][] input = {
-                {0.22805803f, 0.60407318f, 0.91519962f, 0.35643331f, 0.28702669f},
-                {0.11558246f, 0.57658853f, 0.47569648f, 0.02271072f, 0.24709974f}};
-        float[][] expected = {{0.13459886f, 0.43918941f, 0.75030122f, 0.22784227f, 0.17593207f},
-                {0.06310892f, 0.41389921f, 0.32481722f, 0.01156111f, 0.14766179f}};
-        for (TestSession.Mode tfMode : tfModes)
-            try (TestSession session = TestSession.createTestSession(tfMode)) {
-                Ops tf = session.getTF();
-                GeLU instance = new GeLU(tf, true);
-                Operand<TFloat32> result = instance.call(tf.constant(input));
-                session.evaluate(tf.constant(expected), result);
-            }
-    }
+  /** Test of GeLU call method */
+  @Test
+  public void testCallFloatApproximate() {
+    float[][] input = {
+      {0.22805803f, 0.60407318f, 0.91519962f, 0.35643331f, 0.28702669f},
+      {0.11558246f, 0.57658853f, 0.47569648f, 0.02271072f, 0.24709974f}
+    };
+    float[][] expected = {
+      {0.13459886f, 0.43918941f, 0.75030122f, 0.22784227f, 0.17593207f},
+      {0.06310892f, 0.41389921f, 0.32481722f, 0.01156111f, 0.14766179f}
+    };
+    for (TestSession.Mode tfMode : tfModes)
+      try (TestSession session = TestSession.createTestSession(tfMode)) {
+        Ops tf = session.getTF();
+        GeLU instance = new GeLU(tf, true);
+        Operand<TFloat32> result = instance.call(tf.constant(input));
+        session.evaluate(tf.constant(expected), result);
+      }
+  }
 
-    /** Test of GeLU call method */
-    @Test
-    public void testCallDoubleApproximate() {
-        double[][] input = {
-                {0.22805803, 0.60407318, 0.91519962, 0.35643331, 0.28702669},
-                {0.11558246, 0.57658853, 0.47569648, 0.02271072, 0.24709974}};
-        double[][] expected = {{0.13459886, 0.43918941, 0.75030122, 0.22784227, 0.17593207},
-                {0.06310892, 0.41389921, 0.32481722, 0.01156111, 0.14766179}};
-        //for (TestSession.Mode tfMode : tfModes)
-            try (TestSession session = TestSession.createTestSession(TestSession.Mode.GRAPH)) {
-                Ops tf = session.getTF();
-                GeLU instance = new GeLU(tf, true);
-                Operand<TFloat64> result = instance.call(tf.constant(input));
-                session.print(result);
-                session.evaluate(tf.constant(expected), result);
-            }
+  /** Test of GeLU call method */
+  @Test
+  public void testCallDoubleApproximate() {
+    double[][] input = {
+      {0.22805803, 0.60407318, 0.91519962, 0.35643331, 0.28702669},
+      {0.11558246, 0.57658853, 0.47569648, 0.02271072, 0.24709974}
+    };
+    double[][] expected = {
+      {0.13459886, 0.43918941, 0.75030122, 0.22784227, 0.17593207},
+      {0.06310892, 0.41389921, 0.32481722, 0.01156111, 0.14766179}
+    };
+    // for (TestSession.Mode tfMode : tfModes)
+    try (TestSession session = TestSession.createTestSession(TestSession.Mode.GRAPH)) {
+      Ops tf = session.getTF();
+      GeLU instance = new GeLU(tf, true);
+      Operand<TFloat64> result = instance.call(tf.constant(input));
+      session.evaluate(tf.constant(expected), result);
     }
-
-
-}
\ No newline at end of file
+  }
+}
diff --git a/tensorflow-framework/src/test/java/org/tensorflow/framework/activations/HardSigmoidTest.java b/tensorflow-framework/src/test/java/org/tensorflow/framework/activations/HardSigmoidTest.java
index 6e8b00dba3e..02b8e565d66 100644
--- a/tensorflow-framework/src/test/java/org/tensorflow/framework/activations/HardSigmoidTest.java
+++ b/tensorflow-framework/src/test/java/org/tensorflow/framework/activations/HardSigmoidTest.java
@@ -78,7 +78,6 @@ public void testCallInt() {
         Ops tf = session.getTF();
         HardSigmoid instance = new HardSigmoid(tf);
         Operand<TInt32> result = instance.call(tf.constant(input));
-        session.print(result);
         int[] expected = {0, 0, 0, 0, 0};
         session.evaluate(expected, result);
       }
diff --git a/tensorflow-framework/src/test/java/org/tensorflow/framework/initializers/ConstantTest.java b/tensorflow-framework/src/test/java/org/tensorflow/framework/initializers/ConstantTest.java
index 812ef4a1ee3..1233637241e 100644
--- a/tensorflow-framework/src/test/java/org/tensorflow/framework/initializers/ConstantTest.java
+++ b/tensorflow-framework/src/test/java/org/tensorflow/framework/initializers/ConstantTest.java
@@ -51,7 +51,7 @@ public void testCallUInt() {
       try (TestSession session = TestSession.createTestSession(tfMode)) {
         Ops tf = session.getTF();
         Shape shape = Shape.of(2, 2);
-        Constant instance = new Constant(tf, 0xf);
+        Constant instance = new Constant(tf, tf.constant(0xf));
         Operand<TUint8> operand = instance.call(tf.constant(shape), TUint8.class);
         session.evaluate(expected, operand);
       }
@@ -67,7 +67,7 @@ public void testCallInt() {
       try (TestSession session = TestSession.createTestSession(tfMode)) {
         Ops tf = session.getTF();
         Shape shape = Shape.of(2, 2);
-        Constant instance = new Constant(tf, 0xf);
+        Constant instance = new Constant(tf, tf.constant(0xf));
         Operand<TInt32> operand = instance.call(tf.constant(shape), TInt32.class);
         session.evaluate(expected, operand);
       }
@@ -83,7 +83,7 @@ public void testCallLong() {
       try (TestSession session = TestSession.createTestSession(tfMode)) {
         Ops tf = session.getTF();
         Shape shape = Shape.of(2, 2);
-        Constant instance = new Constant(tf, 0xffL);
+        Constant instance = new Constant(tf, tf.constant(0xffL));
         Operand<TInt64> operand = instance.call(tf.constant(shape), TInt64.class);
         session.evaluate(expected, operand);
       }
@@ -97,7 +97,7 @@ public void testCallFloat() {
       try (TestSession session = TestSession.createTestSession(tfMode)) {
         Ops tf = session.getTF();
         Shape shape = Shape.of(2, 2);
-        Constant instance = new Constant(tf, 12.F);
+        Constant instance = new Constant(tf, tf.constant(12.f));
         Operand<TFloat32> operand = instance.call(tf.constant(shape), TFloat32.class);
         session.evaluate(expected, operand);
       }
@@ -112,7 +112,7 @@ public void testCallDouble() {
         Ops tf = session.getTF();
         Shape shape = Shape.of(2, 2);
 
-        Constant instance = new Constant(tf, 11.);
+        Constant instance = new Constant(tf, tf.constant(11.));
         Operand<TFloat64> operand = instance.call(tf.constant(shape), TFloat64.class);
         session.evaluate(expected, operand);
       }
@@ -129,9 +129,8 @@ public void testCallString() {
               Ops tf = session.getTF();
               Shape shape = Shape.of(2, 2);
 
-              Constant instance = new Constant(tf, 22);
+              Constant instance = new Constant(tf, tf.constant(22));
               instance.call(tf.constant(shape), TString.class);
-              fail("IllegalArgumentException  should have been thrown for TString");
             }
           });
   }
@@ -145,7 +144,7 @@ public void testCallBool() {
         Shape shape = Shape.of(2, 2);
         Boolean[] expected = {true, true, true, true};
 
-        Constant instance = new Constant(tf, true);
+        Constant instance = new Constant(tf, tf.constant(true));
         Operand<TBool> operand = instance.call(tf.constant(shape), TBool.class);
         session.evaluate(expected, operand);
       }
@@ -158,7 +157,7 @@ public void testReproducible() {
         Ops tf = session.getTF();
         Shape shape = Shape.of(2, 2);
 
-        Constant instance = new Constant(tf, 11.);
+        Constant instance = new Constant(tf, tf.constant(11.));
         Operand<TFloat64> operand1 = instance.call(tf.constant(shape), TFloat64.class);
         Operand<TFloat64> operand2 = instance.call(tf.constant(shape), TFloat64.class);
         session.evaluate(operand1, operand2);
diff --git a/tensorflow-framework/src/test/java/org/tensorflow/framework/initializers/ZerosTest.java b/tensorflow-framework/src/test/java/org/tensorflow/framework/initializers/ZerosTest.java
index 9e80abb61b7..f81df29dae1 100644
--- a/tensorflow-framework/src/test/java/org/tensorflow/framework/initializers/ZerosTest.java
+++ b/tensorflow-framework/src/test/java/org/tensorflow/framework/initializers/ZerosTest.java
@@ -21,6 +21,8 @@
 import org.tensorflow.op.Ops;
 import org.tensorflow.types.*;
 
+import static org.junit.jupiter.api.Assertions.assertThrows;
+
 /** Test the Zeros initializer */
 public class ZerosTest {
 
@@ -115,14 +117,18 @@ public void testCallDouble() {
   @Test
   public void testCallString() {
     for (TestSession.Mode tfMode : tfModes)
-      try (TestSession session = TestSession.createTestSession(tfMode)) {
-        Ops tf = session.getTF();
-        Shape shape = Shape.of(2, 2);
-
-        Zeros instance = new Zeros(tf);
-        Operand<TString> operand = instance.call(tf.constant(shape), TString.class);
-        session.evaluateString(operand, String::isEmpty);
-      }
+      assertThrows(
+          java.lang.IllegalArgumentException.class,
+          () -> {
+            try (TestSession session = TestSession.createTestSession(tfMode)) {
+              Ops tf = session.getTF();
+              Shape shape = Shape.of(2, 2);
+
+              Zeros instance = new Zeros(tf);
+              Operand<TString> operand = instance.call(tf.constant(shape), TString.class);
+              session.evaluateString(operand, String::isEmpty);
+            }
+          });
   }
 
   /** Test of call method, of class Zeros. */

From 478b78aa7715e202be059a86ea8e8d82bce64f9d Mon Sep 17 00:00:00 2001
From: Jim Clarke <JimClarke5@me.com>
Date: Sat, 6 Mar 2021 18:32:23 -0500
Subject: [PATCH 07/13] Add check for scalar.

---
 .../framework/initializers/Constant.java       |  4 ++++
 .../framework/initializers/ConstantTest.java   | 18 +++++++++++++++++-
 2 files changed, 21 insertions(+), 1 deletion(-)

diff --git a/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Constant.java b/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Constant.java
index f20ff694af8..f7aae0dccf1 100644
--- a/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Constant.java
+++ b/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Constant.java
@@ -47,9 +47,13 @@ public class Constant extends BaseInitializer<TType> {
    *
    * @param tf the TensorFlow Ops
    * @param value the value used for the constant.
+   * @throws IllegalArgumentException if value is not a scalar.
    */
   public Constant(Ops tf, Operand<? extends TType> value) {
     super(tf);
+    if (!value.shape().isScalar()) {
+      throw new IllegalArgumentException("value must be scalar, got shape : " + value.shape());
+    }
     this.value = value;
   }
 
diff --git a/tensorflow-framework/src/test/java/org/tensorflow/framework/initializers/ConstantTest.java b/tensorflow-framework/src/test/java/org/tensorflow/framework/initializers/ConstantTest.java
index 1233637241e..25cec75252c 100644
--- a/tensorflow-framework/src/test/java/org/tensorflow/framework/initializers/ConstantTest.java
+++ b/tensorflow-framework/src/test/java/org/tensorflow/framework/initializers/ConstantTest.java
@@ -22,7 +22,6 @@
 import org.tensorflow.types.*;
 
 import static org.junit.jupiter.api.Assertions.assertThrows;
-import static org.junit.jupiter.api.Assertions.fail;
 
 /** Test the Constant initializer */
 public class ConstantTest {
@@ -135,6 +134,23 @@ public void testCallString() {
           });
   }
 
+  /** Test of call method, of class Constant. */
+  @Test
+  public void testCallNonScalar() {
+    for (TestSession.Mode tfMode : tfModes)
+      assertThrows(
+          java.lang.IllegalArgumentException.class,
+          () -> {
+            try (TestSession session = TestSession.createTestSession(tfMode)) {
+              Ops tf = session.getTF();
+              Shape shape = Shape.of(2, 2);
+
+              Constant instance = new Constant(tf, tf.constant(new int[] {1, 2}));
+              instance.call(tf.constant(shape), TInt32.class);
+            }
+          });
+  }
+
   /** Test of call method, of class Constant. */
   @Test
   public void testCallBool() {

From f53fa0888d2c9b027092b96962541080c1db266c Mon Sep 17 00:00:00 2001
From: Jim Clarke <JimClarke5@me.com>
Date: Sat, 6 Mar 2021 19:10:06 -0500
Subject: [PATCH 08/13] Change to accept TString value.

---
 .../framework/initializers/Constant.java      |  5 +--
 .../framework/initializers/ConstantTest.java  | 37 +++++++++++++++++--
 2 files changed, 34 insertions(+), 8 deletions(-)

diff --git a/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Constant.java b/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Constant.java
index f7aae0dccf1..b14c7fe332e 100644
--- a/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Constant.java
+++ b/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Constant.java
@@ -68,10 +68,7 @@ public Constant(Ops tf, Operand<? extends TType> value) {
    */
   @Override
   public <U extends TType> Operand<U> call(Operand<TInt64> dims, Class<U> type) {
-    if (!TNumber.class.isAssignableFrom(type) && type != TBool.class) {
-      throw new IllegalArgumentException(
-          "Tensor type must be numeric or boolean: " + type.getSimpleName());
-    }
+
     return tf.fill(dims, cast(tf, value, type));
   }
 }
diff --git a/tensorflow-framework/src/test/java/org/tensorflow/framework/initializers/ConstantTest.java b/tensorflow-framework/src/test/java/org/tensorflow/framework/initializers/ConstantTest.java
index 25cec75252c..a81db689191 100644
--- a/tensorflow-framework/src/test/java/org/tensorflow/framework/initializers/ConstantTest.java
+++ b/tensorflow-framework/src/test/java/org/tensorflow/framework/initializers/ConstantTest.java
@@ -120,16 +120,45 @@ public void testCallDouble() {
   /** Test of call method, of class Constant. */
   @Test
   public void testCallString() {
+    for (TestSession.Mode tfMode : tfModes)
+      try (TestSession session = TestSession.createTestSession(tfMode)) {
+        Ops tf = session.getTF();
+        Shape shape = Shape.of(2, 2);
+        String[][] expected = {
+          {"Java Test", "Java Test"},
+          {"Java Test", "Java Test"},
+        };
+        // There is no tf.constant(String[][]).
+        Operand<TString> expectedOp =
+            org.tensorflow.op.core.Constant.tensorOf(tf.scope(), expected);
+
+        Constant instance = new Constant(tf, tf.constant("Java Test"));
+        Operand<TString> result = instance.call(tf.constant(shape), TString.class);
+        session.evaluate(expectedOp, result);
+      }
+  }
+
+  /** Test of call method, of class Constant. */
+  @Test
+  public void testCallStringInvalidDateTYpe() {
     for (TestSession.Mode tfMode : tfModes)
       assertThrows(
-          java.lang.IllegalArgumentException.class,
+          org.tensorflow.exceptions.TFUnimplementedException.class,
           () -> {
             try (TestSession session = TestSession.createTestSession(tfMode)) {
               Ops tf = session.getTF();
               Shape shape = Shape.of(2, 2);
-
-              Constant instance = new Constant(tf, tf.constant(22));
-              instance.call(tf.constant(shape), TString.class);
+              String[][] expected = {
+                {"Java Test", "Java Test"},
+                {"Java Test", "Java Test"},
+              };
+              // There is no tf.constant(String[][]).
+              Operand<TString> expectedOp =
+                  org.tensorflow.op.core.Constant.tensorOf(tf.scope(), expected);
+
+              Constant instance = new Constant(tf, tf.constant("Java Test"));
+              Operand<TInt32> result = instance.call(tf.constant(shape), TInt32.class);
+              session.evaluate(expectedOp, result);
             }
           });
   }

From 79594da9901606c7972c7afe93d87ec297a7521b Mon Sep 17 00:00:00 2001
From: Jim Clarke <JimClarke5@me.com>
Date: Tue, 9 Mar 2021 15:29:17 -0500
Subject: [PATCH 09/13] Fix GeLU equations with separate Operands

---
 .../framework/activations/GeLU.java           | 26 +++++++------------
 1 file changed, 10 insertions(+), 16 deletions(-)

diff --git a/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/GeLU.java b/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/GeLU.java
index 6b27dd4c12a..abddbd512cf 100644
--- a/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/GeLU.java
+++ b/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/GeLU.java
@@ -70,6 +70,9 @@ public GeLU(Ops tf, boolean approximate) {
   @Override
   public <U extends TFloating> Operand<U> call(Operand<U> input) {
 
+    Operand<U> point5 = cast(tf, tf.constant(0.5), input.type());
+    Operand<U> one = cast(tf, tf.constant(1.0), input.type());
+
     if (approximate) {
       /*
               coeff = math_ops.cast(0.044715, features.dtype)
@@ -78,8 +81,9 @@ public <U extends TFloating> Operand<U> call(Operand<U> input) {
                               (features + coeff * math_ops.pow(features, 3))))
       */
       Operand<U> coeff = cast(tf, tf.constant(0.044715), input.type());
-      Operand<U> point5 = cast(tf, tf.constant(0.5), input.type());
-      Operand<U> one = cast(tf, tf.constant(1.0), input.type());
+      // sqrt(2.0 / PI)
+      Operand<U> sqrtTwoDivPI = cast(tf, tf.constant(0.7978845608028654), input.type());
+      Operand<U> three = cast(tf, tf.constant(3), input.type());
 
       return tf.math.mul(
           point5,
@@ -89,13 +93,9 @@ public <U extends TFloating> Operand<U> call(Operand<U> input) {
                   one,
                   tf.math.tanh(
                       tf.math.mul(
-                          // sqrt(2.0 / PI)
-                          cast(tf, tf.constant(0.7978845608028654), input.type()),
+                          sqrtTwoDivPI,
                           tf.math.add(
-                              input,
-                              tf.math.mul(
-                                  coeff,
-                                  tf.math.pow(input, cast(tf, tf.constant(3), input.type()))) // mul
+                              input, tf.math.mul(coeff, tf.math.pow(input, three)) // mul
                               ) // add
                           ) // mul
                       ) // tanh
@@ -108,15 +108,9 @@ public <U extends TFloating> Operand<U> call(Operand<U> input) {
       return 0.5 * features * (1.0 + math_ops.erf(
         features / math_ops.cast(1.4142135623730951, features.dtype)))
        */
+      Operand<U> sqrtTwo = cast(tf, tf.constant(1.4142135623730951), input.type());
       return tf.math.mul(
-          cast(tf, tf.constant(0.5), input.type()),
-          tf.math.mul(
-              input,
-              tf.math.add(
-                  cast(tf, tf.constant(1), input.type()),
-                  tf.math.erf(
-                      tf.math.div(
-                          input, cast(tf, tf.constant(1.4142135623730951), input.type()))))));
+          point5, tf.math.mul(input, tf.math.add(one, tf.math.erf(tf.math.div(input, sqrtTwo)))));
     }
   }
 }

From 112c74023154f0d6eeed95a3fc6eb9a3703e19e3 Mon Sep 17 00:00:00 2001
From: Jim Clarke <JimClarke5@me.com>
Date: Tue, 9 Mar 2021 15:30:10 -0500
Subject: [PATCH 10/13] Fix Constant to handle TString properly  added
 additional test case for invalid conversion of TString to TInt32

---
 .../org/tensorflow/framework/initializers/Constant.java     | 1 -
 .../org/tensorflow/framework/initializers/ConstantTest.java | 6 ++++--
 2 files changed, 4 insertions(+), 3 deletions(-)

diff --git a/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Constant.java b/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Constant.java
index b14c7fe332e..2efb59b8723 100644
--- a/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Constant.java
+++ b/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Constant.java
@@ -64,7 +64,6 @@ public Constant(Ops tf, Operand<? extends TType> value) {
    * @param type the data type of tensor
    * @param <U> The data Type for initializer operation
    * @return An operand for the initialization.
-   * @throws IllegalArgumentException if the data type is not a TNumber or TBool
    */
   @Override
   public <U extends TType> Operand<U> call(Operand<TInt64> dims, Class<U> type) {
diff --git a/tensorflow-framework/src/test/java/org/tensorflow/framework/initializers/ConstantTest.java b/tensorflow-framework/src/test/java/org/tensorflow/framework/initializers/ConstantTest.java
index a81db689191..45c7392c06e 100644
--- a/tensorflow-framework/src/test/java/org/tensorflow/framework/initializers/ConstantTest.java
+++ b/tensorflow-framework/src/test/java/org/tensorflow/framework/initializers/ConstantTest.java
@@ -21,6 +21,8 @@
 import org.tensorflow.op.Ops;
 import org.tensorflow.types.*;
 
+import java.util.Arrays;
+
 import static org.junit.jupiter.api.Assertions.assertThrows;
 
 /** Test the Constant initializer */
@@ -140,7 +142,7 @@ public void testCallString() {
 
   /** Test of call method, of class Constant. */
   @Test
-  public void testCallStringInvalidDateTYpe() {
+  public void testCallStringInvalidDataType() {
     for (TestSession.Mode tfMode : tfModes)
       assertThrows(
           org.tensorflow.exceptions.TFUnimplementedException.class,
@@ -158,7 +160,7 @@ public void testCallStringInvalidDateTYpe() {
 
               Constant instance = new Constant(tf, tf.constant("Java Test"));
               Operand<TInt32> result = instance.call(tf.constant(shape), TInt32.class);
-              session.evaluate(expectedOp, result);
+              session.run(result); // this will cause the exception to be thrown.
             }
           });
   }

From 61e620620a24985425f2a2fb2f3ea45265088a8f Mon Sep 17 00:00:00 2001
From: Jim Clarke <JimClarke5@me.com>
Date: Tue, 9 Mar 2021 15:30:39 -0500
Subject: [PATCH 11/13] Added Stddev check for not less than 0.

---
 .../framework/initializers/RandomNormal.java  |  4 +++
 .../initializers/RandomNormalTest.java        | 31 +++++++++++--------
 2 files changed, 22 insertions(+), 13 deletions(-)

diff --git a/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/RandomNormal.java b/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/RandomNormal.java
index 789942f52ce..2d71d02e1c1 100644
--- a/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/RandomNormal.java
+++ b/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/RandomNormal.java
@@ -75,9 +75,13 @@ public RandomNormal(Ops tf, double mean, long seed) {
    * @param stddev Standard deviation of the random values to generate.
    * @param seed the seed for random number generation. An initializer created with a given seed
    *     will always produce the same random tensor for a given shape and data type.
+   * @throws IllegalArgumentException if standard deviation is less than 0.
    */
   public RandomNormal(Ops tf, double mean, double stddev, long seed) {
     super(tf);
+    if(stddev < 0) {
+      throw new IllegalArgumentException("Standard deviation (stddev) cannot be less than 0, got " + stddev);
+    }
     this.mean = mean;
     this.stddev = stddev;
     this.seed = seed;
diff --git a/tensorflow-framework/src/test/java/org/tensorflow/framework/initializers/RandomNormalTest.java b/tensorflow-framework/src/test/java/org/tensorflow/framework/initializers/RandomNormalTest.java
index 7f9e5d841f0..268bd515daf 100644
--- a/tensorflow-framework/src/test/java/org/tensorflow/framework/initializers/RandomNormalTest.java
+++ b/tensorflow-framework/src/test/java/org/tensorflow/framework/initializers/RandomNormalTest.java
@@ -14,7 +14,7 @@
 =======================================================================*/
 package org.tensorflow.framework.initializers;
 
-import org.junit.jupiter.api.*;
+import org.junit.jupiter.api.Test;
 import org.tensorflow.Operand;
 import org.tensorflow.framework.utils.TestSession;
 import org.tensorflow.ndarray.Shape;
@@ -22,6 +22,8 @@
 import org.tensorflow.types.TFloat32;
 import org.tensorflow.types.TFloat64;
 
+import static org.junit.jupiter.api.Assertions.assertThrows;
+
 /** Test the RandomNormal initializer */
 public class RandomNormalTest {
 
@@ -32,18 +34,6 @@ public class RandomNormalTest {
 
   public RandomNormalTest() {}
 
-  @BeforeAll
-  public static void setUpClass() {}
-
-  @AfterAll
-  public static void tearDownClass() {}
-
-  @BeforeEach
-  public void setUp() {}
-
-  @AfterEach
-  public void tearDown() {}
-
   /** Test of call method, of class RandomNormal. */
   @Test
   public void testCalltestSoftmaxFloat() {
@@ -86,4 +76,19 @@ public void testReproducible() {
         session.evaluate(operand1, operand2);
       }
   }
+
+  @Test
+  public void testInvalidStdDev() {
+    for (TestSession.Mode tfMode : tfModes)
+      assertThrows(
+          IllegalArgumentException.class,
+          () -> {
+            try (TestSession session = TestSession.createTestSession(tfMode)) {
+              Ops tf = session.getTF();
+              Shape shape = Shape.of(2, 2);
+
+              RandomNormal instance = new RandomNormal(tf, MEAN_VALUE, -2.5, SEED);
+            }
+          });
+  }
 }

From 3b4b6077f3a36ef8efe8b6deab354fcb022fa480 Mon Sep 17 00:00:00 2001
From: Jim Clarke <JimClarke5@me.com>
Date: Tue, 9 Mar 2021 15:31:59 -0500
Subject: [PATCH 12/13] Fix fix fill to cast the 1 to the approriate type
 before the fill

---
 .../main/java/org/tensorflow/framework/initializers/Ones.java   | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Ones.java b/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Ones.java
index e2ad49c07f9..5cf8c7a8033 100644
--- a/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Ones.java
+++ b/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Ones.java
@@ -71,6 +71,6 @@ public <U extends TType> Operand<U> call(Operand<TInt64> dims, Class<U> type) {
           "Tensor type must be numeric or boolean: " + type.getSimpleName());
     }
 
-    return cast(tf, tf.fill(dims, tf.constant(1)), type);
+    return tf.fill(dims, cast(tf, tf.constant(1), type));
   }
 }

From 98df654a9558d0d41ab35766073f7ad97038d556 Mon Sep 17 00:00:00 2001
From: Jim Clarke <JimClarke5@me.com>
Date: Tue, 9 Mar 2021 15:42:16 -0500
Subject: [PATCH 13/13] Code reformat

---
 .../org/tensorflow/framework/activations/Exponential.java    | 1 -
 .../java/org/tensorflow/framework/initializers/Constant.java | 2 --
 .../org/tensorflow/framework/initializers/RandomNormal.java  | 5 +++--
 .../org/tensorflow/framework/initializers/ConstantTest.java  | 2 --
 .../tensorflow/framework/initializers/RandomNormalTest.java  | 1 -
 5 files changed, 3 insertions(+), 8 deletions(-)

diff --git a/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Exponential.java b/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Exponential.java
index 79857dd3f7b..512f96713aa 100644
--- a/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Exponential.java
+++ b/tensorflow-framework/src/main/java/org/tensorflow/framework/activations/Exponential.java
@@ -52,7 +52,6 @@ public Exponential(Ops tf) {
   @Override
   public <U extends TFloating> Operand<U> call(Operand<U> input) {
 
-
     return tf.math.exp(input);
   }
 }
diff --git a/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Constant.java b/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Constant.java
index 2efb59b8723..3f2ebe58cb4 100644
--- a/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Constant.java
+++ b/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/Constant.java
@@ -16,9 +16,7 @@
 
 import org.tensorflow.Operand;
 import org.tensorflow.op.Ops;
-import org.tensorflow.types.TBool;
 import org.tensorflow.types.TInt64;
-import org.tensorflow.types.family.TNumber;
 import org.tensorflow.types.family.TType;
 
 import static org.tensorflow.framework.utils.CastHelper.cast;
diff --git a/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/RandomNormal.java b/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/RandomNormal.java
index 2d71d02e1c1..6b90ed3985d 100644
--- a/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/RandomNormal.java
+++ b/tensorflow-framework/src/main/java/org/tensorflow/framework/initializers/RandomNormal.java
@@ -79,8 +79,9 @@ public RandomNormal(Ops tf, double mean, long seed) {
    */
   public RandomNormal(Ops tf, double mean, double stddev, long seed) {
     super(tf);
-    if(stddev < 0) {
-      throw new IllegalArgumentException("Standard deviation (stddev) cannot be less than 0, got " + stddev);
+    if (stddev < 0) {
+      throw new IllegalArgumentException(
+          "Standard deviation (stddev) cannot be less than 0, got " + stddev);
     }
     this.mean = mean;
     this.stddev = stddev;
diff --git a/tensorflow-framework/src/test/java/org/tensorflow/framework/initializers/ConstantTest.java b/tensorflow-framework/src/test/java/org/tensorflow/framework/initializers/ConstantTest.java
index 45c7392c06e..a66b013a0c2 100644
--- a/tensorflow-framework/src/test/java/org/tensorflow/framework/initializers/ConstantTest.java
+++ b/tensorflow-framework/src/test/java/org/tensorflow/framework/initializers/ConstantTest.java
@@ -21,8 +21,6 @@
 import org.tensorflow.op.Ops;
 import org.tensorflow.types.*;
 
-import java.util.Arrays;
-
 import static org.junit.jupiter.api.Assertions.assertThrows;
 
 /** Test the Constant initializer */
diff --git a/tensorflow-framework/src/test/java/org/tensorflow/framework/initializers/RandomNormalTest.java b/tensorflow-framework/src/test/java/org/tensorflow/framework/initializers/RandomNormalTest.java
index 268bd515daf..5c0811fbc5a 100644
--- a/tensorflow-framework/src/test/java/org/tensorflow/framework/initializers/RandomNormalTest.java
+++ b/tensorflow-framework/src/test/java/org/tensorflow/framework/initializers/RandomNormalTest.java
@@ -85,7 +85,6 @@ public void testInvalidStdDev() {
           () -> {
             try (TestSession session = TestSession.createTestSession(tfMode)) {
               Ops tf = session.getTF();
-              Shape shape = Shape.of(2, 2);
 
               RandomNormal instance = new RandomNormal(tf, MEAN_VALUE, -2.5, SEED);
             }