Merge branch 'master' into bob/nbench_db

doc/sphinx/source/python_api/_autosummary/ngraph.ops.rst

@@ -32,6 +32,7 @@ ngraph.ops
       cosh
       divide
       dot
+      elu
       equal
       exp
       floor

python/ngraph/__init__.py

@@ -45,6 +45,7 @@ from ngraph.ops import cos
 from ngraph.ops import cosh
 from ngraph.ops import divide
 from ngraph.ops import dot
+from ngraph.ops import elu
 from ngraph.ops import equal
 from ngraph.ops import exp
 from ngraph.ops import floor

python/ngraph/impl/op/__init__.py

@@ -69,6 +69,7 @@ from _pyngraph.op import Cos
 from _pyngraph.op import Cosh
 from _pyngraph.op import Divide
 from _pyngraph.op import Dot
+from _pyngraph.op import Elu
 from _pyngraph.op import Equal
 from _pyngraph.op import Exp
 from _pyngraph.op import Floor

python/ngraph/ops.py

@@ -22,7 +22,7 @@ from ngraph.impl import AxisSet, AxisVector, Coordinate, CoordinateDiff, Functio
 
 from ngraph.impl.op import Abs, Acos, Add, And, Asin, ArgMax, ArgMin, Atan, AvgPool, \
     BatchNormTraining, BatchNormInference, Broadcast, Ceiling, Concat, Constant, Convert, \
-    Convolution, ConvolutionBackpropData, Cos, Cosh, Divide, Dot, Equal, Exp, Floor, \
+    Convolution, ConvolutionBackpropData, Cos, Cosh, Divide, Dot, Elu, Equal, Exp, Floor, \
     GetOutputElement, Greater, GreaterEq, Less, LessEq, Log, LRN, Max, Maximum, MaxPool, \
     Min, Minimum, Multiply, Negative, Not, NotEqual, OneHot, Or, Pad, Parameter, Product, \
     Power, Relu, ReplaceSlice, Reshape, Reverse, Select, Sign, Sin, Sinh, Slice, Softmax, \
@@ -35,7 +35,7 @@ from ngraph.utils.decorators import nameable_op, binary_op, unary_op
 from ngraph.utils.input_validation import assert_list_of_ints
 from ngraph.utils.reduction import get_reduction_axes
 from ngraph.utils.types import NumericType, NumericData, TensorShape, make_constant_node, \
-    NodeInput, ScalarData
+    NodeInput, ScalarData, as_node
 from ngraph.utils.types import get_element_type
 
 
@@ -60,6 +60,24 @@ def constant(value, dtype=None, name=None):  # type: (NumericData, NumericType,
     return make_constant_node(value, dtype)
 
 
+@nameable_op
+def elu(data, alpha, name=None):  # type: (NodeInput, NodeInput, str) -> Node
+    """Perform Exponential Linear Unit operation element-wise on data from input node.
+
+    Computes exponential linear: alpha * (exp(data) - 1) if < 0, data otherwise.
+
+    For more information refer to:
+    `Fast and Accurate Deep Network Learning by Exponential Linear Units (ELUs)
+    <http://arxiv.org/abs/1511.07289>`_
+
+    :param data: Input tensor. One of: input node, array or scalar.
+    :param alpha: Multiplier for negative values. One of: input node or scalar value.
+    :param name: Optional output node name.
+    :return: The new node performing an ELU operation on its input data element-wise.
+    """
+    return Elu(as_node(data), as_node(alpha))
+
+
 # Unary ops
 @unary_op
 def absolute(node, name=None):  # type: (NodeInput, str) -> Node

python/pyngraph/ops/elu.cpp

+//*****************************************************************************
+// Copyright 2017-2019 Intel Corporation
+//
+// 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.
+//*****************************************************************************
+
+#include <pybind11/pybind11.h>
+#include <pybind11/stl.h>
+
+#include "ngraph/op/fused/elu.hpp"
+#include "pyngraph/ops/elu.hpp"
+
+namespace py = pybind11;
+
+void regclass_pyngraph_op_Elu(py::module m)
+{
+    py::class_<ngraph::op::Elu, std::shared_ptr<ngraph::op::Elu>, ngraph::op::Op> elu(m, "Elu");
+    elu.doc() = "ngraph.impl.op.Elu wraps ngraph::op::Elu";
+    elu.def(py::init<const std::shared_ptr<ngraph::Node>&, const std::shared_ptr<ngraph::Node>&>());
+}

python/pyngraph/ops/elu.hpp

+//*****************************************************************************
+// Copyright 2017-2019 Intel Corporation
+//
+// 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.
+//*****************************************************************************
+
+#pragma once
+
+#include <pybind11/pybind11.h>
+
+namespace py = pybind11;
+
+void regclass_pyngraph_op_Elu(py::module m);

python/pyngraph/ops/fused/elu.cpp

+//*****************************************************************************
+// Copyright 2017-2019 Intel Corporation
+//
+// 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.
+//*****************************************************************************
+
+#include <pybind11/pybind11.h>
+#include <pybind11/stl.h>
+
+#include "ngraph/op/fused/elu.hpp"
+#include "pyngraph/ops/fused/elu.hpp"
+
+namespace py = pybind11;
+
+void regclass_pyngraph_op_Elu(py::module m)
+{
+    py::class_<ngraph::op::Elu, std::shared_ptr<ngraph::op::Elu>, ngraph::op::Op> elu(m, "Elu");
+    elu.doc() = "ngraph.impl.op.Elu wraps ngraph::op::Elu";
+    elu.def(py::init<const std::shared_ptr<ngraph::Node>&, const std::shared_ptr<ngraph::Node>&>());
+}

python/pyngraph/ops/regmodule_pyngraph_op.cpp

@@ -49,6 +49,7 @@ void regmodule_pyngraph_op(py::module m_op)
     regclass_pyngraph_op_Cosh(m_op);
     regclass_pyngraph_op_Divide(m_op);
     regclass_pyngraph_op_Dot(m_op);
+    regclass_pyngraph_op_Elu(m_op);
     regclass_pyngraph_op_Equal(m_op);
     regclass_pyngraph_op_Exp(m_op);
     regclass_pyngraph_op_Floor(m_op);

python/pyngraph/ops/regmodule_pyngraph_op.hpp

@@ -39,6 +39,7 @@
 #include "pyngraph/ops/cosh.hpp"
 #include "pyngraph/ops/divide.hpp"
 #include "pyngraph/ops/dot.hpp"
+#include "pyngraph/ops/elu.hpp"
 #include "pyngraph/ops/equal.hpp"
 #include "pyngraph/ops/exp.hpp"
 #include "pyngraph/ops/floor.hpp"

python/setup.py

@@ -179,6 +179,7 @@ sources = [
     'pyngraph/ops/ceiling.cpp',
     'pyngraph/ops/divide.cpp',
     'pyngraph/ops/dot.cpp',
+    'pyngraph/ops/elu.cpp',
     'pyngraph/ops/equal.cpp',
     'pyngraph/ops/exp.cpp',
     'pyngraph/ops/floor.cpp',

python/test/ngraph/test_ops_fused.py

+# ******************************************************************************
+# Copyright 2017-2019 Intel Corporation
+#
+# 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.
+# ******************************************************************************
+import numpy as np
+
+import ngraph as ng
+from test.ngraph.util import get_runtime
+
+
+def test_elu_operator():
+    runtime = get_runtime()
+
+    data_shape = [2, 2]
+    alpha_shape = [2]
+    parameter_data = ng.parameter(data_shape, name='Data', dtype=np.float32)
+    parameter_alpha = ng.parameter(alpha_shape, name='Alpha', dtype=np.float32)
+
+    model = ng.elu(parameter_data, parameter_alpha)
+    computation = runtime.computation(model, parameter_data, parameter_alpha)
+
+    value_data = np.array([[-5, 1], [-2, 3]], dtype=np.float32)
+    value_alpha = np.array([3, 3], dtype=np.float32)
+
+    result = computation(value_data, value_alpha)
+    expected = np.array([[-2.9797862, 1.], [-2.5939941, 3.]], dtype=np.float32)
+    assert np.allclose(result, expected)
+
+
+def test_elu_operator_with_scalar_and_array():
+    runtime = get_runtime()
+
+    data_value = np.array([[-5, 1], [-2, 3]], dtype=np.float32)
+    alpha_value = np.float32(3)
+
+    model = ng.elu(data_value, alpha_value)
+    computation = runtime.computation(model)
+
+    result = computation()
+    expected = np.array([[-2.9797862, 1.], [-2.5939941, 3.]], dtype=np.float32)
+    assert np.allclose(result, expected)
+
+
+def test_elu_operator_with_scalar():
+    runtime = get_runtime()
+
+    data_value = np.array([[-5, 1], [-2, 3]], dtype=np.float32)
+    alpha_value = np.float32(3)
+
+    data_shape = [2, 2]
+    parameter_data = ng.parameter(data_shape, name='Data', dtype=np.float32)
+
+    model = ng.elu(parameter_data, alpha_value)
+    computation = runtime.computation(model, parameter_data)
+
+    result = computation(data_value)
+    expected = np.array([[-2.9797862, 1.], [-2.5939941, 3.]], dtype=np.float32)
+    assert np.allclose(result, expected)

python/test/ngraph/util.py

@@ -16,9 +16,8 @@
 
 import numpy as np
 import ngraph as ng
-
-from string import ascii_uppercase
-
+from ngraph.utils.types import NumericData
+from typing import Any, Callable, List
 import test
 
 
@@ -32,10 +31,14 @@ def get_runtime():
 
 
 def run_op_node(input_data, op_fun, *args):
+    # type: (NumericData, Callable, *Any) -> List[NumericData]
     """Run computation on node performing `op_fun`.
 
     `op_fun` has to accept a node as an argument.
 
+    This function converts passed raw input data to nGraph Constant Node and that form is passed
+    to `op_fun`.
+
     :param input_data: The input data for performed computation.
     :param op_fun: The function handler for operation we want to carry out.
     :param args: The arguments passed to operation we want to carry out.
@@ -45,14 +48,8 @@ def run_op_node(input_data, op_fun, *args):
     comp_args = []
     op_fun_args = []
     comp_inputs = []
-    for idx, data in enumerate(input_data):
-        if np.isscalar(data):
-            op_fun_args.append(ng.constant(data, _get_numpy_dtype(data)))
-        else:
-            node = ng.parameter(data.shape, name=ascii_uppercase[idx], dtype=data.dtype)
-            op_fun_args.append(node)
-            comp_args.append(node)
-            comp_inputs.append(data)
+    for data in input_data:
+        op_fun_args.append(ng.constant(data, _get_numpy_dtype(data)))
     op_fun_args.extend(args)
     node = op_fun(*op_fun_args)
     computation = runtime.computation(node, *comp_args)
@@ -60,10 +57,15 @@ def run_op_node(input_data, op_fun, *args):
 
 
 def run_op_numeric_data(input_data, op_fun, *args):
+    # type: (NumericData, Callable, *Any) -> List[NumericData]
     """Run computation on node performing `op_fun`.
 
     `op_fun` has to accept a scalar or an array.
 
+    This function passess input data AS IS. This mean that in case they're a scalar (integral,
+    or floating point value) or a NumPy's ndarray object they will be automatically converted
+    to nGraph's Constant Nodes.
+
     :param input_data: The input data for performed computation.
     :param op_fun: The function handler for operation we want to carry out.
     :param args: The arguments passed to operation we want to carry out.

src/contrib/mlir/lowerer.cpp

@@ -350,6 +350,7 @@ namespace
         }
         return callBackFuncPtr;
     }
+
     // NGDialect converters
     Type NGraphTypeConverter::convertType(Type type)
     {
@@ -576,7 +577,6 @@ namespace
         // Create Value for result, and extract type info.
         Value* result = m_pass.buildOutputDefs(op, rewriter)[0];
         NGRAPH_CHECK(result, "Unexpected null result in ConcatOp");
-        auto resultTy = result->getType().cast<MemRefType>();
 
         // Create view to write into result.
         MemRefView vRes(result);
@@ -590,7 +590,6 @@ namespace
         for (auto& operand : operands)
         {
             NGRAPH_CHECK(operand, "Unexpected null operand in ConcatOp");
-            auto operandTy = result->getType().cast<MemRefType>();
 
             // Assuming rank = r, and the concatenation axis is A where A<r, we'll be creating
             // loops of this form:

src/contrib/mlir/pass/mlir_subgraph_extraction.cpp

@@ -74,7 +74,6 @@ void MLIRSubgraphExtractionPass::MLIRSubgraph::merge(MLIRSubgraph& sg2)
 
     // Associate nodes of second sub-graph to first one
     auto sg_nodes = sg2.get_nodes();
-    auto& node_map = m_pass.m_node_to_graph;
     for (auto node : sg_nodes)
     {
         NGRAPH_DEBUG << *node;
@@ -112,7 +111,6 @@ bool MLIRSubgraphExtractionPass::run_on_function(std::shared_ptr<Function> func)
     for (auto op : func->get_ordered_ops())
     {
         NodeVector inputs;
-        int first_graph_id = -1;
         std::unordered_set<int> subgraph_ids;
         // unsupported ops, skip
         if (!is_supported_mlir_op(op))

src/ngraph/check.hpp

@@ -160,5 +160,5 @@ namespace ngraph
 /// \brief Macro to signal a code path that is unreachable in a successful execution. It's
 /// implemented with NGRAPH_CHECK macro.
 /// \param ... Additional error message that should describe why that execution path is unreachable.
-/// \throws ::ngrap::CheckFailure if the macro is executed.
+/// \throws ::ngraph::CheckFailure if the macro is executed.
 #define NGRAPH_UNREACHABLE(...) NGRAPH_CHECK(false, "Unreachable: ", ##__VA_ARGS__)

src/ngraph/node.hpp

@@ -214,7 +214,7 @@ namespace ngraph
         virtual bool is_constant() const;
         virtual bool is_null() const { return false; }
         virtual bool is_op() const { return false; }
-        virtual bool is_commutative() { return false; }
+        virtual bool is_commutative() const { return false; }
         virtual bool is_dynamic() const;
         virtual bool has_state() const { return false; }
         size_t get_instance_id() const { return m_instance_id; }

src/ngraph/op/add.cpp

@@ -49,7 +49,7 @@ void op::Add::generate_adjoints(autodiff::Adjoints& adjoints, const NodeVector&
     adjoints.add_delta(y, delta);
 }
 
-shared_ptr<Node> ngraph::operator+(const shared_ptr<Node>& arg0, const shared_ptr<Node>& arg1)
+shared_ptr<Node> ngraph::operator+(const Output<Node>& arg0, const Output<Node>& arg1)
 {
     return make_shared<op::Add>(arg0, arg1);
 }

src/ngraph/op/add.hpp

@@ -51,13 +51,12 @@ namespace ngraph
 
             std::shared_ptr<Node> copy_with_new_args(const NodeVector& new_args) const override;
 
+            virtual bool is_commutative() const override { return true; }
         protected:
             virtual void generate_adjoints(autodiff::Adjoints& adjoints,
                                            const NodeVector& deltas) override;
-            virtual bool is_commutative() override { return true; }
         };
     }
 
-    std::shared_ptr<ngraph::Node> operator+(const std::shared_ptr<ngraph::Node>& arg0,
-                                            const std::shared_ptr<ngraph::Node>& arg1);
+    std::shared_ptr<Node> operator+(const Output<Node>& arg0, const Output<Node>& arg1);
 }

src/ngraph/op/and.hpp

@@ -51,8 +51,7 @@ namespace ngraph
 
             std::shared_ptr<Node> copy_with_new_args(const NodeVector& new_args) const override;
 
-        protected:
-            virtual bool is_commutative() override { return true; }
+            virtual bool is_commutative() const override { return true; }
         };
     }
 }

src/ngraph/op/batch_norm.cpp

@@ -22,12 +22,15 @@
 #include "ngraph/op/get_output_element.hpp"
 #include "ngraph/validation_util.hpp"
 
-const std::string ngraph::op::BatchNormTraining::type_name{"BatchNormTraining"};
+using namespace std;
+using namespace ngraph;
 
-ngraph::op::BatchNormTraining::BatchNormTraining(Output<ngraph::Node> input,
-                                                 Output<ngraph::Node> gamma,
-                                                 Output<ngraph::Node> beta,
-                                                 double epsilon)
+const string op::BatchNormTraining::type_name{"BatchNormTraining"};
+
+op::BatchNormTraining::BatchNormTraining(const Output<Node>& input,
+                                         const Output<Node>& gamma,
+                                         const Output<Node>& beta,
+                                         double epsilon)
     : Op({gamma, beta, input})
     , m_epsilon(epsilon)
 {
@@ -35,17 +38,17 @@ ngraph::op::BatchNormTraining::BatchNormTraining(Output<ngraph::Node> input,
 }
 
 // DEPRECATED
-ngraph::op::BatchNormTraining::BatchNormTraining(double eps,
-                                                 Output<ngraph::Node> gamma,
-                                                 Output<ngraph::Node> beta,
-                                                 Output<ngraph::Node> input)
+op::BatchNormTraining::BatchNormTraining(double eps,
+                                         const Output<Node>& gamma,
+                                         const Output<Node>& beta,
+                                         const Output<Node>& input)
     : Op({gamma, beta, input})
     , m_epsilon(eps)
 {
     constructor_validate_and_infer_types();
 }
 
-void ngraph::op::BatchNormTraining::validate_and_infer_types()
+void op::BatchNormTraining::validate_and_infer_types()
 {
     element::Type result_et;
     PartialShape result_batch_shape;
@@ -66,16 +69,15 @@ void ngraph::op::BatchNormTraining::validate_and_infer_types()
     set_output_type(2, result_et, result_channel_shape);
 }
 
-std::shared_ptr<ngraph::Node>
-    ngraph::op::BatchNormTraining::copy_with_new_args(const NodeVector& new_args) const
+std::shared_ptr<Node> op::BatchNormTraining::copy_with_new_args(const NodeVector& new_args) const
 {
     check_new_args_count(this, new_args);
     return std::make_shared<BatchNormTraining>(
         new_args.at(2), new_args.at(0), new_args.at(1), m_epsilon);
 }
 
-void ngraph::op::BatchNormTraining::generate_adjoints(autodiff::Adjoints& adjoints,
-                                                      const NodeVector& deltas)
+void op::BatchNormTraining::generate_adjoints(autodiff::Adjoints& adjoints,
+                                              const NodeVector& deltas)
 {
     auto gamma = input(0).get_source_output();
     auto beta = input(1).get_source_output();
@@ -102,14 +104,14 @@ void ngraph::op::BatchNormTraining::generate_adjoints(autodiff::Adjoints& adjoin
     adjoints.add_delta(beta, dbeta);
 }
 
-const std::string ngraph::op::BatchNormInference::type_name{"BatchNormInference"};
+const string op::BatchNormInference::type_name{"BatchNormInference"};
 
-ngraph::op::BatchNormInference::BatchNormInference(Output<ngraph::Node> input,
-                                                   Output<ngraph::Node> gamma,
-                                                   Output<ngraph::Node> beta,
-                                                   Output<ngraph::Node> mean,
-                                                   Output<ngraph::Node> variance,
-                                                   double epsilon)
+op::BatchNormInference::BatchNormInference(const Output<Node>& input,
+                                           const Output<Node>& gamma,
+                                           const Output<Node>& beta,
+                                           const Output<Node>& mean,
+                                           const Output<Node>& variance,
+                                           double epsilon)
     : Op({gamma, beta, input, mean, variance})
     , m_epsilon(epsilon)
 {
@@ -117,19 +119,19 @@ ngraph::op::BatchNormInference::BatchNormInference(Output<ngraph::Node> input,
 }
 
 // DEPRECATED
-ngraph::op::BatchNormInference::BatchNormInference(double eps,
-                                                   Output<ngraph::Node> gamma,
-                                                   Output<ngraph::Node> beta,
-                                                   Output<ngraph::Node> input,
-                                                   Output<ngraph::Node> mean,
-                                                   Output<ngraph::Node> variance)
+op::BatchNormInference::BatchNormInference(double eps,
+                                           const Output<Node>& gamma,
+                                           const Output<Node>& beta,
+                                           const Output<Node>& input,
+                                           const Output<Node>& mean,
+                                           const Output<Node>& variance)
     : Op({gamma, beta, input, mean, variance})
     , m_epsilon(eps)
 {
     constructor_validate_and_infer_types();
 }
 
-void ngraph::op::BatchNormInference::validate_and_infer_types()
+void op::BatchNormInference::validate_and_infer_types()
 {
     element::Type result_et;
     PartialShape result_batch_shape;
@@ -152,23 +154,22 @@ void ngraph::op::BatchNormInference::validate_and_infer_types()
     set_output_type(0, result_et, result_batch_shape);
 }
 
-std::shared_ptr<ngraph::Node>
-    ngraph::op::BatchNormInference::copy_with_new_args(const NodeVector& new_args) const
+std::shared_ptr<Node> op::BatchNormInference::copy_with_new_args(const NodeVector& new_args) const
 {
     check_new_args_count(this, new_args);
     return std::make_shared<BatchNormInference>(
         new_args.at(2), new_args.at(0), new_args.at(1), new_args.at(3), new_args.at(4), m_epsilon);
 }
 
-const std::string ngraph::op::BatchNormTrainingBackprop::type_name{"BatchNormTrainingBackprop"};
+const string op::BatchNormTrainingBackprop::type_name{"BatchNormTrainingBackprop"};
 
-ngraph::op::BatchNormTrainingBackprop::BatchNormTrainingBackprop(Output<ngraph::Node> input,
-                                                                 Output<ngraph::Node> gamma,
-                                                                 Output<ngraph::Node> beta,
-                                                                 Output<ngraph::Node> mean,
-                                                                 Output<ngraph::Node> variance,
-                                                                 Output<ngraph::Node> delta,
-                                                                 double epsilon)
+op::BatchNormTrainingBackprop::BatchNormTrainingBackprop(const Output<Node>& input,
+                                                         const Output<Node>& gamma,
+                                                         const Output<Node>& beta,
+                                                         const Output<Node>& mean,
+                                                         const Output<Node>& variance,
+                                                         const Output<Node>& delta,
+                                                         double epsilon)
     : Op({gamma, beta, input, mean, variance, delta})
     , m_epsilon(epsilon)
 
@@ -177,13 +178,13 @@ ngraph::op::BatchNormTrainingBackprop::BatchNormTrainingBackprop(Output<ngraph::
     constructor_validate_and_infer_types();
 }
 
-ngraph::op::BatchNormTrainingBackprop::BatchNormTrainingBackprop(double epsilon,
-                                                                 Output<ngraph::Node> gamma,
-                                                                 Output<ngraph::Node> beta,
-                                                                 Output<ngraph::Node> input,
-                                                                 Output<ngraph::Node> mean,
-                                                                 Output<ngraph::Node> variance,
-                                                                 Output<ngraph::Node> delta)
+op::BatchNormTrainingBackprop::BatchNormTrainingBackprop(double epsilon,
+                                                         const Output<Node>& gamma,
+                                                         const Output<Node>& beta,
+                                                         const Output<Node>& input,
+                                                         const Output<Node>& mean,
+                                                         const Output<Node>& variance,
+                                                         const Output<Node>& delta)
     : Op({gamma, beta, input, mean, variance, delta})
     , m_epsilon(epsilon)
 
@@ -192,7 +193,7 @@ ngraph::op::BatchNormTrainingBackprop::BatchNormTrainingBackprop(double epsilon,
     constructor_validate_and_infer_types();
 }
 
-void ngraph::op::BatchNormTrainingBackprop::validate_and_infer_types()
+void op::BatchNormTrainingBackprop::validate_and_infer_types()
 {
     PartialShape input_and_delta_shape{get_input_partial_shape(INPUT_DATA)};
 
@@ -239,8 +240,8 @@ void ngraph::op::BatchNormTrainingBackprop::validate_and_infer_types()
     set_output_type(2, result_et, result_channel_shape);
 }
 
-std::shared_ptr<ngraph::Node>
-    ngraph::op::BatchNormTrainingBackprop::copy_with_new_args(const NodeVector& new_args) const
+std::shared_ptr<Node>
+    op::BatchNormTrainingBackprop::copy_with_new_args(const NodeVector& new_args) const
 {
     check_new_args_count(this, new_args);
     return std::make_shared<op::BatchNormTrainingBackprop>(new_args.at(2),

src/ngraph/op/batch_norm.hpp

@@ -39,9 +39,9 @@ namespace ngraph
             /// \param gamma gamma scaling for normalized value. [C]
             /// \param beta bias added to the scaled normalized value [C]
             /// \param epsilon Avoids divsion by 0 if input has 0 variance
-            BatchNormTraining(Output<Node> input,
-                              Output<Node> gamma,
-                              Output<Node> beta,
+            BatchNormTraining(const Output<Node>& input,
+                              const Output<Node>& gamma,
+                              const Output<Node>& beta,
                               double epsilon);
 
             NGRAPH_DEPRECATED_DOC
@@ -66,9 +66,9 @@ namespace ngraph
             ///   output[2]: shall have rank 1, with the same span as input's channel axis.
             NGRAPH_DEPRECATED("Use another constructor")
             BatchNormTraining(double eps,
-                              Output<Node> gamma,
-                              Output<Node> beta,
-                              Output<Node> input);
+                              const Output<Node>& gamma,
+                              const Output<Node>& beta,
+                              const Output<Node>& input);
 
             void validate_and_infer_types() override;
 
@@ -101,11 +101,11 @@ namespace ngraph
             /// \param mean value for mean normalization [C]
             /// \param variance value for variance normalization [C]
             /// \param epsilon Avoids divsion by 0 if input has 0 variance
-            BatchNormInference(Output<ngraph::Node> input,
-                               Output<ngraph::Node> gamma,
-                               Output<ngraph::Node> beta,
-                               Output<ngraph::Node> mean,
-                               Output<ngraph::Node> variance,
+            BatchNormInference(const Output<Node>& input,
+                               const Output<Node>& gamma,
+                               const Output<Node>& beta,
+                               const Output<Node>& mean,
+                               const Output<Node>& variance,
                                double epsilon);
 
             NGRAPH_DEPRECATED_DOC
@@ -128,11 +128,11 @@ namespace ngraph
             ///   output:   shall have the same shape as 'input'.
             NGRAPH_DEPRECATED("Use another constructor")
             BatchNormInference(double eps,
-                               Output<ngraph::Node> gamma,
-                               Output<ngraph::Node> beta,
-                               Output<ngraph::Node> input,
-                               Output<ngraph::Node> mean,
-                               Output<ngraph::Node> variance);
+                               const Output<Node>& gamma,
+                               const Output<Node>& beta,
+                               const Output<Node>& input,
+                               const Output<Node>& mean,
+                               const Output<Node>& variance);
 
             void validate_and_infer_types() override;
 
@@ -165,24 +165,23 @@ namespace ngraph
             static const std::string type_name;
             const std::string& description() const override { return type_name; }
             BatchNormTrainingBackprop() = default;
-            BatchNormTrainingBackprop(Output<Node> input,
-                                      Output<Node> gamma,
-                                      Output<Node> beta,
-                                      Output<Node> mean,
-                                      Output<Node> variance,
-                                      Output<Node> delta,
+            BatchNormTrainingBackprop(const Output<Node>& input,
+                                      const Output<Node>& gamma,
+                                      const Output<Node>& beta,
+                                      const Output<Node>& mean,
+                                      const Output<Node>& variance,
+                                      const Output<Node>& delta,
                                       double epsilon);
 
             NGRAPH_DEPRECATED_DOC
             NGRAPH_DEPRECATED("Use another constructor")
             BatchNormTrainingBackprop(double epsilon,
-                                      Output<Node> gamma,
-                                      Output<Node> beta,
-                                      Output<Node> input,
-
-                                      Output<Node> mean,
-                                      Output<Node> variance,
-                                      Output<Node> delta);
+                                      const Output<Node>& gamma,
+                                      const Output<Node>& beta,
+                                      const Output<Node>& input,
+                                      const Output<Node>& mean,
+                                      const Output<Node>& variance,
+                                      const Output<Node>& delta);
 
             void validate_and_infer_types() override;
 

src/ngraph/op/divide.cpp

@@ -64,7 +64,7 @@ void op::Divide::generate_adjoints(autodiff::Adjoints& adjoints, const NodeVecto
     adjoints.add_delta(y, -delta * shared_from_this() / y);
 }
 
-shared_ptr<Node> ngraph::operator/(const Output<Node> arg0, const Output<Node> arg1)
+shared_ptr<Node> ngraph::operator/(const Output<Node>& arg0, const Output<Node>& arg1)
 {
     return make_shared<op::Divide>(arg0, arg1);
 }

src/ngraph/op/divide.hpp

@@ -64,6 +64,5 @@ namespace ngraph
         };
     }
 
-    std::shared_ptr<ngraph::Node> operator/(const Output<ngraph::Node> arg0,
-                                            const Output<ngraph::Node> arg1);
+    std::shared_ptr<Node> operator/(const Output<Node>& arg0, const Output<Node>& arg1);
 }

src/ngraph/op/dot.hpp

@@ -58,7 +58,7 @@ namespace ngraph
             void validate_and_infer_types() override;
 
             size_t get_reduction_axes_count() const { return m_reduction_axes_count; }
-            void get_reduction_axes_count(size_t reduction_axes_count)
+            void set_reduction_axes_count(size_t reduction_axes_count)
             {
                 m_reduction_axes_count = reduction_axes_count;
             }

src/ngraph/op/equal.hpp

@@ -56,6 +56,8 @@ namespace ngraph
 
             virtual std::shared_ptr<Node>
                 copy_with_new_args(const NodeVector& new_args) const override;
+
+            virtual bool is_commutative() const override { return true; }
         };
     }
 }

src/ngraph/op/fused/elu.cpp

@@ -39,7 +39,7 @@ NodeVector op::Elu::decompose_op() const
     auto data = get_argument(0);
     auto alpha_node = get_argument(1);
 
-    alpha_node = ngraph::op::make_broadcast_node(alpha_node, data->get_shape());
+    alpha_node = ngraph::op::numpy_style_broadcast(alpha_node, data->get_shape());
 
     shared_ptr<ngraph::Node> zero_node =
         builder::make_constant(data->get_element_type(), data->get_shape(), 0);

src/ngraph/op/gather.cpp

@@ -23,6 +23,8 @@ using namespace ngraph;
 static int PARAMS = 0;
 static int INDICES = 1;
 
+const string op::Gather::type_name{"Gather"};
+
 shared_ptr<Node> op::Gather::copy_with_new_args(const NodeVector& new_args) const
 {
     check_new_args_count(this, new_args);

src/ngraph/op/gather.hpp

@@ -26,13 +26,15 @@ namespace ngraph
         class Gather : public Op
         {
         public:
+            NGRAPH_API
+            static const std::string type_name;
+            const std::string& description() const override { return type_name; }
+            Gather() = default;
             /// \param params The tensor from which slices are gathered
             /// \param indices Index tensor: Data type must be `element::i32` or `element::i64`
             /// \param axis Axis in params to gather
-            Gather(const std::shared_ptr<Node>& params,
-                   const std::shared_ptr<Node>& indices,
-                   size_t axis = 0)
-                : Op("Gather", check_single_output_args({params, indices}))
+            Gather(const Output<Node>& params, const Output<Node>& indices, size_t axis = 0)
+                : Op({params, indices})
                 , m_axis(axis)
             {
                 constructor_validate_and_infer_types();
@@ -46,6 +48,7 @@ namespace ngraph
             }
 
             size_t get_axis() const { return m_axis; }
+            void set_axis(size_t axis) { m_axis = axis; }
             virtual std::shared_ptr<Node>
                 copy_with_new_args(const NodeVector& new_args) const override;
 

src/ngraph/op/gather_nd.cpp

@@ -23,6 +23,8 @@ using namespace ngraph;
 static int PARAMS = 0;
 static int INDICES = 1;
 
+const string op::GatherND::type_name{"GatherND"};
+
 shared_ptr<Node> op::GatherND::copy_with_new_args(const NodeVector& new_args) const
 {
     check_new_args_count(this, new_args);

src/ngraph/op/gather_nd.hpp

@@ -26,10 +26,14 @@ namespace ngraph
         class GatherND : public Op
         {
         public:
+            NGRAPH_API
+            static const std::string type_name;
+            const std::string& description() const override { return type_name; }
+            GatherND() = default;
             /// \param params The tensor from which slices are gathered
             /// \param indices Index tensor: Data type must be `element::i32` or `element::i64`
-            GatherND(const std::shared_ptr<Node>& params, const std::shared_ptr<Node>& indices)
-                : Op("GatherND", check_single_output_args({params, indices}))
+            GatherND(const Output<Node>& params, const Output<Node>& indices)
+                : Op({params, indices})
             {
                 constructor_validate_and_infer_types();
             }

src/ngraph/op/greater.cpp

@@ -19,10 +19,12 @@
 using namespace std;
 using namespace ngraph;
 
-op::Greater::Greater(const shared_ptr<Node>& arg0,
-                     const shared_ptr<Node>& arg1,
+const string op::Greater::type_name{"Greater"};
+
+op::Greater::Greater(const Output<Node>& arg0,
+                     const Output<Node>& arg1,
                      const AutoBroadcastSpec& autob)
-    : BinaryElementwiseComparison("Greater", arg0, arg1, autob)
+    : BinaryElementwiseComparison(arg0, arg1, autob)
 {
     constructor_validate_and_infer_types();
 }

src/ngraph/op/greater.hpp

@@ -26,13 +26,18 @@ namespace ngraph
         class Greater : public util::BinaryElementwiseComparison
         {
         public:
+            NGRAPH_API
+            static const std::string type_name;
+            const std::string& description() const override { return type_name; }
+            /// \brief Constructs a greater-than operation.
+            Greater() = default;
             /// \brief Constructs a greater-than operation.
             ///
             /// \param arg0 Node that produces the first input tensor.
             /// \param arg1 Node that produces the second input tensor.
             /// \param autob Auto broadcast specification
-            Greater(const std::shared_ptr<Node>& arg0,
-                    const std::shared_ptr<Node>& arg1,
+            Greater(const Output<Node>& arg0,
+                    const Output<Node>& arg1,
                     const AutoBroadcastSpec& autob = AutoBroadcastSpec());
 
             virtual std::shared_ptr<Node>

src/ngraph/op/greater_eq.cpp

@@ -19,10 +19,12 @@
 using namespace std;
 using namespace ngraph;
 
-op::GreaterEq::GreaterEq(const shared_ptr<Node>& arg0,
-                         const shared_ptr<Node>& arg1,
+const string op::GreaterEq::type_name{"GreaterEq"};
+
+op::GreaterEq::GreaterEq(const Output<Node>& arg0,
+                         const Output<Node>& arg1,
                          const AutoBroadcastSpec& autob)
-    : BinaryElementwiseComparison("GreaterEq", arg0, arg1, autob)
+    : BinaryElementwiseComparison(arg0, arg1, autob)
 {
     constructor_validate_and_infer_types();
 }

src/ngraph/op/greater_eq.hpp

@@ -26,13 +26,18 @@ namespace ngraph
         class GreaterEq : public util::BinaryElementwiseComparison
         {
         public:
+            NGRAPH_API
+            static const std::string type_name;
+            const std::string& description() const override { return type_name; }
+            /// \brief Constructs a greater-than-or-equal operation.
+            GreaterEq() = default;
             /// \brief Constructs a greater-than-or-equal operation.
             ///
             /// \param arg0 Node that produces the first input tensor.
             /// \param arg1 Node that produces the second input tensor.
             /// \param autob Auto broadcast specification
-            GreaterEq(const std::shared_ptr<Node>& arg0,
-                      const std::shared_ptr<Node>& arg1,
+            GreaterEq(const Output<Node>& arg0,
+                      const Output<Node>& arg1,
                       const AutoBroadcastSpec& autob = AutoBroadcastSpec());
 
             virtual std::shared_ptr<Node>

src/ngraph/op/less.cpp

@@ -19,10 +19,10 @@
 using namespace std;
 using namespace ngraph;
 
-op::Less::Less(const shared_ptr<Node>& arg0,
-               const shared_ptr<Node>& arg1,
-               const AutoBroadcastSpec& autob)
-    : BinaryElementwiseComparison("Less", arg0, arg1, autob)
+const string op::Less::type_name{"Less"};
+
+op::Less::Less(const Output<Node>& arg0, const Output<Node>& arg1, const AutoBroadcastSpec& autob)
+    : BinaryElementwiseComparison(arg0, arg1, autob)
 {
     constructor_validate_and_infer_types();
 }

src/ngraph/op/less.hpp

@@ -26,13 +26,18 @@ namespace ngraph
         class Less : public util::BinaryElementwiseComparison
         {
         public:
+            NGRAPH_API
+            static const std::string type_name;
+            const std::string& description() const override { return type_name; }
+            /// \brief Constructs a less-than operation.
+            Less() = default;
             /// \brief Constructs a less-than operation.
             ///
             /// \param arg0 Node that produces the first input tensor.
             /// \param arg1 Node that produces the second input tensor.
             /// \param autob Auto broadcast specification
-            Less(const std::shared_ptr<Node>& arg0,
-                 const std::shared_ptr<Node>& arg1,
+            Less(const Output<Node>& arg0,
+                 const Output<Node>& arg1,
                  const AutoBroadcastSpec& autob = AutoBroadcastSpec());
 
             virtual std::shared_ptr<Node>

src/ngraph/op/less_eq.cpp

@@ -19,10 +19,12 @@
 using namespace std;
 using namespace ngraph;
 
-op::LessEq::LessEq(const shared_ptr<Node>& arg0,
-                   const shared_ptr<Node>& arg1,
+const string op::LessEq::type_name{"LessEq"};
+
+op::LessEq::LessEq(const Output<Node>& arg0,
+                   const Output<Node>& arg1,
                    const AutoBroadcastSpec& autob)
-    : BinaryElementwiseComparison("LessEq", arg0, arg1, autob)
+    : BinaryElementwiseComparison(arg0, arg1, autob)
 {
     constructor_validate_and_infer_types();
 }

src/ngraph/op/less_eq.hpp

@@ -26,13 +26,18 @@ namespace ngraph
         class LessEq : public util::BinaryElementwiseComparison
         {
         public:
+            NGRAPH_API
+            static const std::string type_name;
+            const std::string& description() const override { return type_name; }
+            /// \brief Constructs a less-than-or-equal operation.
+            LessEq() = default;
             /// \brief Constructs a less-than-or-equal operation.
             ///
             /// \param arg0 Node that produces the first input tensor.
             /// \param arg1 Node that produces the second input tensor.
             /// \param autob Auto broadcast specification
-            LessEq(const std::shared_ptr<Node>& arg0,
-                   const std::shared_ptr<Node>& arg1,
+            LessEq(const Output<Node>& arg0,
+                   const Output<Node>& arg1,
                    const AutoBroadcastSpec& autob = AutoBroadcastSpec());
 
             virtual std::shared_ptr<Node>

src/ngraph/op/log.cpp

@@ -20,8 +20,10 @@
 using namespace std;
 using namespace ngraph;
 
-op::Log::Log(const shared_ptr<Node>& arg)
-    : UnaryElementwiseArithmetic("Log", arg)
+const string op::Log::type_name{"Log"};
+
+op::Log::Log(const Output<Node>& arg)
+    : UnaryElementwiseArithmetic(arg)
 {
     constructor_validate_and_infer_types();
 }

src/ngraph/op/log.hpp

@@ -26,10 +26,15 @@ namespace ngraph
         class Log : public util::UnaryElementwiseArithmetic
         {
         public:
+            NGRAPH_API
+            static const std::string type_name;
+            const std::string& description() const override { return type_name; }
+            /// \brief Constructs a natural log operation.
+            Log() = default;
             /// \brief Constructs a natural log operation.
             ///
             /// \param arg Node that produces the input tensor.
-            Log(const std::shared_ptr<Node>& arg);
+            Log(const Output<Node>& arg);
 
             virtual std::shared_ptr<Node>
                 copy_with_new_args(const NodeVector& new_args) const override;

src/ngraph/op/lrn.cpp

@@ -20,12 +20,14 @@
 using namespace std;
 using namespace ngraph;
 
-op::LRN::LRN(const std::shared_ptr<Node>& arg, double alpha, double beta, double bias, size_t nsize)
-    : UnaryElementwiseArithmetic("LRN", arg)
+const string op::LRN::type_name{"LRN"};
+
+op::LRN::LRN(const Output<Node>& arg, double alpha, double beta, double bias, size_t size)
+    : UnaryElementwiseArithmetic(arg)
     , m_alpha(alpha)
     , m_beta(beta)
     , m_bias(bias)
-    , m_size(nsize)
+    , m_size(size)
 {
     constructor_validate_and_infer_types();
 }

src/ngraph/op/lrn.hpp

@@ -38,23 +38,28 @@ namespace ngraph
         class LRN : public util::UnaryElementwiseArithmetic
         {
         public:
+            NGRAPH_API
+            static const std::string type_name;
+            const std::string& description() const override { return type_name; }
+            /// \brief Constructs a LRN operation.
+            LRN() = default;
             /// \brief Constructs a LRN operation.
             ///
             /// \param arg Node that produces the input tensor.
-            LRN(const std::shared_ptr<Node>& arg,
-                double alpha,
-                double beta,
-                double bias,
-                size_t size);
+            LRN(const Output<Node>& arg, double alpha, double beta, double bias, size_t size);
 
             virtual std::shared_ptr<Node>
                 copy_with_new_args(const NodeVector& new_args) const override;
             void validate_and_infer_types() override;
 
             double get_alpha() const { return m_alpha; }
+            void set_alpha(double alpha) { m_alpha = alpha; }
             double get_beta() const { return m_beta; }
+            void set_beta(double beta) { m_beta = beta; }
             double get_bias() const { return m_bias; }
+            void set_bias(double bias) { m_bias = bias; }
             size_t get_nsize() const { return m_size; }
+            void set_nsize(size_t size) { m_size = size; }
         protected:
             virtual void generate_adjoints(autodiff::Adjoints& adjoints,
                                            const NodeVector& deltas) override;

src/ngraph/op/max.cpp

@@ -22,10 +22,6 @@ using namespace ngraph;
 
 const string op::Max::type_name{"Max"};
 
-op::Max::Max()
-{
-}
-
 op::Max::Max(const Output<Node>& arg, const AxisSet& reduction_axes)
     : ArithmeticReduction(arg, reduction_axes)
 {

src/ngraph/op/max.hpp

@@ -30,7 +30,7 @@ namespace ngraph
             static const std::string type_name;
             const std::string& description() const override { return type_name; }
             /// \brief Constructs a "max" reduction operation.
-            Max();
+            Max() = default;
             /// \brief Constructs a max-reduction operation.
             ///
             /// \param arg The tensor to be reduced.

src/ngraph/op/max_pool.cpp

@@ -25,14 +25,16 @@
 using namespace std;
 using namespace ngraph;
 
-op::MaxPool::MaxPool(const shared_ptr<Node>& arg,
+const string op::MaxPool::type_name{"MaxPool"};
+
+op::MaxPool::MaxPool(const Output<Node>& arg,
                      const Shape& window_shape,
                      const Strides& window_movement_strides,
                      const Shape& padding_below,
                      const Shape& padding_above,
                      const PadType& pad_type,
                      bool ceil_mode)
-    : Op("MaxPool", check_single_output_args({arg}))
+    : Op({arg})
     , m_window_shape(window_shape)
     , m_window_movement_strides(window_movement_strides)
     , m_padding_below(padding_below)
@@ -43,7 +45,7 @@ op::MaxPool::MaxPool(const shared_ptr<Node>& arg,
     constructor_validate_and_infer_types();
 }
 
-op::MaxPool::MaxPool(const shared_ptr<Node>& arg,
+op::MaxPool::MaxPool(const Output<Node>& arg,
                      const Shape& window_shape,
                      const Strides& window_movement_strides,
                      const Shape& padding_below,
@@ -54,7 +56,7 @@ op::MaxPool::MaxPool(const shared_ptr<Node>& arg,
 {
 }
 
-op::MaxPool::MaxPool(const shared_ptr<Node>& arg,
+op::MaxPool::MaxPool(const Output<Node>& arg,
                      const Shape& window_shape,
                      const Strides& window_movement_strides,
                      const Shape& padding_below,
@@ -121,14 +123,14 @@ void op::MaxPool::validate_and_infer_types()
                                                   m_ceil_mode));
 }
 
-op::MaxPool::MaxPool(const shared_ptr<Node>& arg,
+op::MaxPool::MaxPool(const Output<Node>& arg,
                      const Shape& window_shape,
                      const Strides& window_movement_strides)
     : MaxPool(arg, window_shape, window_movement_strides, Shape(), Shape())
 {
 }
 
-op::MaxPool::MaxPool(const shared_ptr<Node>& arg, const Shape& window_shape)
+op::MaxPool::MaxPool(const Output<Node>& arg, const Shape& window_shape)
     : MaxPool(arg, window_shape, Strides(), Shape(), Shape())
 {
 }
@@ -145,13 +147,15 @@ shared_ptr<Node> op::MaxPool::copy_with_new_args(const NodeVector& new_args) con
                                 m_ceil_mode);
 }
 
-op::MaxPoolBackprop::MaxPoolBackprop(const shared_ptr<Node>& arg_forward,
-                                     const shared_ptr<Node>& delta,
+const string op::MaxPoolBackprop::type_name{"MaxPoolBackprop"};
+
+op::MaxPoolBackprop::MaxPoolBackprop(const Output<Node>& arg_forward,
+                                     const Output<Node>& delta,
                                      const Shape& window_shape,
                                      const Strides& window_movement_strides,
                                      const Shape& padding_below,
                                      const Shape& padding_above)
-    : Op("MaxPoolBackprop", check_single_output_args({arg_forward, delta}))
+    : Op({arg_forward, delta})
     , m_window_shape(window_shape)
     , m_window_movement_strides(window_movement_strides)
     , m_padding_below(padding_below)
@@ -160,14 +164,14 @@ op::MaxPoolBackprop::MaxPoolBackprop(const shared_ptr<Node>& arg_forward,
     constructor_validate_and_infer_types();
 }
 
-op::MaxPoolBackprop::MaxPoolBackprop(const shared_ptr<Node>& arg_forward,
-                                     const shared_ptr<Node>& delta,
-                                     const shared_ptr<Node>& result_forward,
+op::MaxPoolBackprop::MaxPoolBackprop(const Output<Node>& arg_forward,
+                                     const Output<Node>& delta,
+                                     const Output<Node>& result_forward,
                                      const Shape& window_shape,
                                      const Strides& window_movement_strides,
                                      const Shape& padding_below,
                                      const Shape& padding_above)
-    : Op("MaxPoolBackprop", check_single_output_args({arg_forward, delta, result_forward}))
+    : Op({arg_forward, delta, result_forward})
     , m_window_shape(window_shape)
     , m_window_movement_strides(window_movement_strides)
     , m_padding_below(padding_below)

src/ngraph/op/max_pool.hpp

@@ -28,6 +28,12 @@ namespace ngraph
         class MaxPool : public Op
         {
         public:
+            NGRAPH_API
+            static const std::string type_name;
+            const std::string& description() const override { return type_name; }
+            /// \brief Constructs a batched max pooling operation.
+            MaxPool() = default;
+
             /// \brief Constructs a batched max pooling operation.
             ///
             /// \param arg The node producing the input data batch tensor.
@@ -37,7 +43,7 @@ namespace ngraph
             /// \param padding_above The above-padding shape.
             /// \param pad_type The pad type for automatically computing padding sizes
             /// \param ceil_mode Whether to use ceiling while computing output shape.
-            MaxPool(const std::shared_ptr<Node>& arg,
+            MaxPool(const Output<Node>& arg,
                     const Shape& window_shape,
                     const Strides& window_movement_strides,
                     const Shape& padding_below,
@@ -53,7 +59,7 @@ namespace ngraph
             /// \param padding_below The below-padding shape.
             /// \param padding_above The above-padding shape.
             /// \param pad_type The pad type for automatically computing padding sizes
-            MaxPool(const std::shared_ptr<Node>& arg,
+            MaxPool(const Output<Node>& arg,
                     const Shape& window_shape,
                     const Strides& window_movement_strides,
                     const Shape& padding_below,
@@ -67,7 +73,7 @@ namespace ngraph
             /// \param window_movement_strides The window movement strides.
             /// \param padding_below The below-padding shape.
             /// \param padding_above The above-padding shape.
-            MaxPool(const std::shared_ptr<Node>& arg,
+            MaxPool(const Output<Node>& arg,
                     const Shape& window_shape,
                     const Strides& window_movement_strides,
                     const Shape& padding_below,
@@ -80,7 +86,7 @@ namespace ngraph
             /// \param arg The node producing the input data batch tensor.
             /// \param window_shape The window shape.
             /// \param window_movement_strides The window movement strides.
-            MaxPool(const std::shared_ptr<Node>& arg,
+            MaxPool(const Output<Node>& arg,
                     const Shape& window_shape,
                     const Strides& window_movement_strides);
 
@@ -88,23 +94,32 @@ namespace ngraph
             ///
             /// \param arg The node producing the input data batch tensor.
             /// \param window_shape The window shape.
-            MaxPool(const std::shared_ptr<Node>& arg, const Shape& window_shape);
+            MaxPool(const Output<Node>& arg, const Shape& window_shape);
 
             virtual std::shared_ptr<Node>
                 copy_with_new_args(const NodeVector& new_args) const override;
 
             /// \return The window shape.
             const Shape& get_window_shape() const { return m_window_shape; }
+            void set_window_shape(const Shape& window_shape) { m_window_shape = window_shape; }
             /// \return The window movement strides.
             const Strides& get_window_movement_strides() const { return m_window_movement_strides; }
+            void set_window_movement_strides(const Strides& window_movement_strides)
+            {
+                m_window_movement_strides = window_movement_strides;
+            }
             /// \return The below-padding shape.
             const Shape& get_padding_below() const { return m_padding_below; }
+            void set_padding_below(const Shape& padding_below) { m_padding_below = padding_below; }
             /// \return The above-padding shape.
             const Shape& get_padding_above() const { return m_padding_above; }
+            void set_adding_above(const Shape& padding_above) { m_padding_above = padding_above; }
             /// \return The pad type for pooling.
             const PadType& get_pad_type() const { return m_pad_type; }
+            void set_pad_type(const PadType& pad_type) { m_pad_type = pad_type; }
             /// \return The ceiling mode being used for output shape computations
             bool get_ceil_mode() const { return m_ceil_mode; }
+            void set_ceil_mode(bool ceil_mode) { m_ceil_mode = ceil_mode; }
             /// \return The default value for MaxPool.
             virtual std::shared_ptr<Node> get_default_value() const override
             {
@@ -126,16 +141,21 @@ namespace ngraph
         class MaxPoolBackprop : public Op
         {
         public:
-            MaxPoolBackprop(const std::shared_ptr<Node>& arg_forward,
-                            const std::shared_ptr<Node>& delta,
+            NGRAPH_API
+            static const std::string type_name;
+            const std::string& description() const override { return type_name; }
+            MaxPoolBackprop() = default;
+
+            MaxPoolBackprop(const Output<Node>& arg_forward,
+                            const Output<Node>& delta,
                             const Shape& window_shape,
                             const Strides& window_movement_strides,
                             const Shape& padding_below,
                             const Shape& padding_above);
 
-            MaxPoolBackprop(const std::shared_ptr<Node>& arg_forward,
-                            const std::shared_ptr<Node>& delta,
-                            const std::shared_ptr<Node>& result_forward,
+            MaxPoolBackprop(const Output<Node>& arg_forward,
+                            const Output<Node>& delta,
+                            const Output<Node>& result_forward,
                             const Shape& window_shape,
                             const Strides& window_movement_strides,
                             const Shape& padding_below,
@@ -147,9 +167,16 @@ namespace ngraph
             void validate_and_infer_types() override;
 
             const Shape& get_window_shape() const { return m_window_shape; }
+            void set_window_shape(const Shape& window_shape) { m_window_shape = window_shape; }
             const Strides& get_window_movement_strides() const { return m_window_movement_strides; }
+            void set_window_movement_strides(const Strides& window_movement_strides)
+            {
+                m_window_movement_strides = window_movement_strides;
+            }
             const Shape& get_padding_below() const { return m_padding_below; }
+            void set_padding_below(const Shape& padding_below) { m_padding_below = padding_below; }
             const Shape& get_padding_above() const { return m_padding_above; }
+            void set_padding_above(const Shape& padding_above) { m_padding_above = padding_above; }
         protected:
             Shape m_window_shape;
             Strides m_window_movement_strides;

src/ngraph/op/maximum.cpp

@@ -25,10 +25,12 @@
 using namespace std;
 using namespace ngraph;
 
-op::Maximum::Maximum(const shared_ptr<Node>& arg0,
-                     const shared_ptr<Node>& arg1,
+const string op::Maximum::type_name{"Maximum"};
+
+op::Maximum::Maximum(const Output<Node>& arg0,
+                     const Output<Node>& arg1,
                      const AutoBroadcastSpec& autob)
-    : BinaryElementwiseArithmetic("Maximum", arg0, arg1, autob)
+    : BinaryElementwiseArithmetic(arg0, arg1, autob)
 {
     constructor_validate_and_infer_types();
 }

src/ngraph/op/maximum.hpp

@@ -26,19 +26,24 @@ namespace ngraph
         class Maximum : public util::BinaryElementwiseArithmetic
         {
         public:
+            NGRAPH_API
+            static const std::string type_name;
+            const std::string& description() const override { return type_name; }
+            /// \brief Constructs a maximum operation.
+            Maximum() = default;
             /// \brief Constructs a maximum operation.
             ///
             /// \param arg0 Node that produces the first input tensor.
             /// \param arg1 Node that produces the second input tensor.
             /// \param autob Auto broadcast specification
-            Maximum(const std::shared_ptr<Node>& arg0,
-                    const std::shared_ptr<Node>& arg1,
+            Maximum(const Output<Node>& arg0,
+                    const Output<Node>& arg1,
                     const AutoBroadcastSpec& autob = AutoBroadcastSpec());
 
             virtual std::shared_ptr<Node>
                 copy_with_new_args(const NodeVector& new_args) const override;
 
-            virtual bool is_commutative() override { return true; }
+            virtual bool is_commutative() const override { return true; }
         protected:
             virtual void generate_adjoints(autodiff::Adjoints& adjoints,
                                            const NodeVector& deltas) override;

src/ngraph/op/min.cpp

@@ -22,10 +22,6 @@ using namespace ngraph;
 
 const string op::Min::type_name{"Min"};
 
-op::Min::Min()
-{
-}
-
 op::Min::Min(const Output<Node>& arg, const AxisSet& reduction_axes)
     : ArithmeticReduction(arg, reduction_axes)
 {

src/ngraph/op/min.hpp

@@ -30,7 +30,7 @@ namespace ngraph
             static const std::string type_name;
             const std::string& description() const override { return type_name; }
             /// \brief Constructs a "min" reduction operation.
-            Min();
+            Min() = default;
             /// \brief Constructs a min-reduction operation.
             ///
             /// \param arg The tensor to be reduced.

src/ngraph/op/minimum.cpp

@@ -25,10 +25,12 @@
 using namespace std;
 using namespace ngraph;
 
-op::Minimum::Minimum(const shared_ptr<Node>& arg0,
-                     const shared_ptr<Node>& arg1,
+const string op::Minimum::type_name{"Minimum"};
+
+op::Minimum::Minimum(const Output<Node>& arg0,
+                     const Output<Node>& arg1,
                      const AutoBroadcastSpec& autob)
-    : BinaryElementwiseArithmetic("Minimum", arg0, arg1, autob)
+    : BinaryElementwiseArithmetic(arg0, arg1, autob)
 {
     constructor_validate_and_infer_types();
 }

src/ngraph/op/minimum.hpp

@@ -26,18 +26,24 @@ namespace ngraph
         class Minimum : public util::BinaryElementwiseArithmetic
         {
         public:
+            NGRAPH_API
+            static const std::string type_name;
+            const std::string& description() const override { return type_name; }
+            /// \brief Constructs a minimum operation.
+            Minimum() = default;
             /// \brief Constructs a minimum operation.
             ///
             /// \param arg0 Node that produces the first input tensor.
             /// \param arg1 Node that produces the second input tensor.
             /// \param autob Auto broadcast specification
-            Minimum(const std::shared_ptr<Node>& arg0,
-                    const std::shared_ptr<Node>& arg1,
+            Minimum(const Output<Node>& arg0,
+                    const Output<Node>& arg1,
                     const AutoBroadcastSpec& autob = AutoBroadcastSpec());
 
             virtual std::shared_ptr<Node>
                 copy_with_new_args(const NodeVector& new_args) const override;
 
+            virtual bool is_commutative() const override { return true; }
         protected:
             virtual void generate_adjoints(autodiff::Adjoints& adjoints,
                                            const NodeVector& deltas) override;

src/ngraph/op/multiply.cpp

@@ -19,10 +19,12 @@
 using namespace std;
 using namespace ngraph;
 
-op::Multiply::Multiply(const shared_ptr<Node>& arg0,
-                       const shared_ptr<Node>& arg1,
+const string op::Multiply::type_name{"Multiply"};
+
+op::Multiply::Multiply(const Output<Node>& arg0,
+                       const Output<Node>& arg1,
                        const AutoBroadcastSpec& autob)
-    : BinaryElementwiseArithmetic("Multiply", arg0, arg1, autob)
+    : BinaryElementwiseArithmetic(arg0, arg1, autob)
 {
     constructor_validate_and_infer_types();
 }
@@ -49,7 +51,7 @@ void op::Multiply::generate_adjoints(autodiff::Adjoints& adjoints, const NodeVec
     adjoints.add_delta(y, x * delta);
 }
 
-shared_ptr<Node> ngraph::operator*(const shared_ptr<Node> arg0, const shared_ptr<Node> arg1)
+shared_ptr<Node> ngraph::operator*(const Output<Node>& arg0, const Output<Node>& arg1)
 {
     return make_shared<op::Multiply>(arg0, arg1);
 }

src/ngraph/op/multiply.hpp

@@ -26,25 +26,29 @@ namespace ngraph
         class Multiply : public util::BinaryElementwiseArithmetic
         {
         public:
+            NGRAPH_API
+            static const std::string type_name;
+            const std::string& description() const override { return type_name; }
+            /// \brief Constructs a multiplication operation.
+            Multiply() = default;
             /// \brief Constructs a multiplication operation.
             ///
             /// \param arg0 Node that produces the first input tensor.
             /// \param arg1 Node that produces the second input tensor.
             /// \param autob Auto broadcast specification
-            Multiply(const std::shared_ptr<Node>& arg0,
-                     const std::shared_ptr<Node>& arg1,
+            Multiply(const Output<Node>& arg0,
+                     const Output<Node>& arg1,
                      const AutoBroadcastSpec& autob = AutoBroadcastSpec());
 
             virtual std::shared_ptr<Node>
                 copy_with_new_args(const NodeVector& new_args) const override;
 
+            virtual bool is_commutative() const override { return true; }
         protected:
             virtual void generate_adjoints(autodiff::Adjoints& adjoints,
                                            const NodeVector& deltas) override;
-            virtual bool is_commutative() override { return true; }
         };
     };
 
-    std::shared_ptr<ngraph::Node> operator*(const std::shared_ptr<ngraph::Node> arg0,
-                                            const std::shared_ptr<ngraph::Node> arg1);
+    std::shared_ptr<Node> operator*(const Output<Node>& arg0, const Output<Node>& arg1);
 }

src/ngraph/op/negative.cpp

@@ -19,8 +19,10 @@
 using namespace std;
 using namespace ngraph;
 
-op::Negative::Negative(const shared_ptr<Node>& arg)
-    : UnaryElementwiseArithmetic("Negative", arg)
+const string op::Negative::type_name{"Negative"};
+
+op::Negative::Negative(const Output<Node>& arg)
+    : UnaryElementwiseArithmetic(arg)
 {
     constructor_validate_and_infer_types();
 }
@@ -40,7 +42,7 @@ void op::Negative::generate_adjoints(autodiff::Adjoints& adjoints, const NodeVec
     adjoints.add_delta(x, -delta);
 }
 
-shared_ptr<Node> ngraph::operator-(const shared_ptr<Node> arg0)
+shared_ptr<Node> ngraph::operator-(const Output<Node>& arg0)
 {
     return make_shared<op::Negative>(arg0);
 }

src/ngraph/op/negative.hpp

@@ -26,17 +26,23 @@ namespace ngraph
         class Negative : public util::UnaryElementwiseArithmetic
         {
         public:
+            NGRAPH_API
+            static const std::string type_name;
+            const std::string& description() const override { return type_name; }
+            /// \brief Constructs a negative operation.
+            Negative() = default;
             /// \brief Constructs a negative operation.
             ///
             /// \param arg Node that produces the input tensor.
-            Negative(const std::shared_ptr<Node>& arg);
+            Negative(const Output<Node>& arg);
 
             virtual std::shared_ptr<Node>
                 copy_with_new_args(const NodeVector& new_args) const override;
 
+        protected:
             virtual void generate_adjoints(autodiff::Adjoints& adjoints,
                                            const NodeVector& deltas) override;
         };
     }
-    std::shared_ptr<ngraph::Node> operator-(const std::shared_ptr<ngraph::Node> arg0);
+    std::shared_ptr<Node> operator-(const Output<Node>& arg0);
 }

src/ngraph/op/not.cpp

@@ -20,8 +20,10 @@
 using namespace ngraph;
 using namespace std;
 
-op::Not::Not(const shared_ptr<Node>& arg)
-    : Op("Not", check_single_output_args({arg}))
+const string op::Not::type_name{"Not"};
+
+op::Not::Not(const Output<Node>& arg)
+    : Op({arg})
 {
     constructor_validate_and_infer_types();
 }

src/ngraph/op/not.hpp

@@ -26,10 +26,15 @@ namespace ngraph
         class Not : public Op
         {
         public:
+            NGRAPH_API
+            static const std::string type_name;
+            const std::string& description() const override { return type_name; }
+            /// \brief Constructs a logical negation operation.
+            Not() = default;
             /// \brief Constructs a logical negation operation.
             ///
             /// \param arg Node that produces the input tensor.
-            Not(const std::shared_ptr<Node>& arg);
+            Not(const Output<Node>& arg);
 
             void validate_and_infer_types() override;
 

src/ngraph/op/not_equal.cpp

@@ -19,10 +19,12 @@
 using namespace std;
 using namespace ngraph;
 
-op::NotEqual::NotEqual(const shared_ptr<Node>& arg0,
-                       const shared_ptr<Node>& arg1,
+const string op::NotEqual::type_name{"NotEqual"};
+
+op::NotEqual::NotEqual(const Output<Node>& arg0,
+                       const Output<Node>& arg1,
                        const AutoBroadcastSpec& autob)
-    : BinaryElementwiseComparison("NotEqual", arg0, arg1, autob)
+    : BinaryElementwiseComparison(arg0, arg1, autob)
 {
     constructor_validate_and_infer_types();
 }

src/ngraph/op/not_equal.hpp

@@ -26,17 +26,24 @@ namespace ngraph
         class NotEqual : public util::BinaryElementwiseComparison
         {
         public:
+            NGRAPH_API
+            static const std::string type_name;
+            const std::string& description() const override { return type_name; }
+            /// \brief Constructs a not-equal operation.
+            NotEqual() = default;
             /// \brief Constructs a not-equal operation.
             ///
             /// \param arg0 Node that produces the first input tensor.
             /// \param arg1 Node that produces the second input tensor.
             /// \param autob Auto broadcast specification
-            NotEqual(const std::shared_ptr<Node>& arg0,
-                     const std::shared_ptr<Node>& arg1,
+            NotEqual(const Output<Node>& arg0,
+                     const Output<Node>& arg1,
                      const AutoBroadcastSpec& autob = AutoBroadcastSpec());
 
             virtual std::shared_ptr<Node>
                 copy_with_new_args(const NodeVector& new_args) const override;
+
+            virtual bool is_commutative() const override { return true; }
         };
     }
 }

src/ngraph/op/one_hot.cpp

@@ -20,8 +20,10 @@
 using namespace std;
 using namespace ngraph;
 
-op::OneHot::OneHot(const shared_ptr<Node>& arg, const PartialShape& shape, size_t one_hot_axis)
-    : Op("OneHot", check_single_output_args({arg}))
+const string op::OneHot::type_name{"OneHot"};
+
+op::OneHot::OneHot(const Output<Node>& arg, const PartialShape& shape, size_t one_hot_axis)
+    : Op({arg})
     , m_shape(shape)
     , m_one_hot_axis(one_hot_axis)
 {

src/ngraph/op/one_hot.hpp

@@ -45,14 +45,17 @@ namespace ngraph
         class OneHot : public Op
         {
         public:
+            NGRAPH_API
+            static const std::string type_name;
+            const std::string& description() const override { return type_name; }
+            /// \brief Constructs a one-hot operation.
+            OneHot() = default;
             /// \brief Constructs a one-hot operation.
             ///
             /// \param arg          Node that produces the input tensor to be one-hot encoded.
             /// \param shape        The shape of the output tensor, including the new one-hot axis.
             /// \param one_hot_axis The index within the output shape of the new one-hot axis.
-            OneHot(const std::shared_ptr<Node>& arg,
-                   const PartialShape& shape,
-                   size_t one_hot_axis);
+            OneHot(const Output<Node>& arg, const PartialShape& shape, size_t one_hot_axis);
 
             virtual std::shared_ptr<Node>
                 copy_with_new_args(const NodeVector& new_args) const override;
@@ -60,6 +63,7 @@ namespace ngraph
 
             /// \return The index of the one-hot axis.
             size_t get_one_hot_axis() const { return m_one_hot_axis; }
+            void set_one_hot_axis(size_t one_hot_axis) { m_one_hot_axis = one_hot_axis; }
         protected:
             PartialShape m_shape;
             size_t m_one_hot_axis;

src/ngraph/op/or.cpp

@@ -19,10 +19,10 @@
 using namespace std;
 using namespace ngraph;
 
-op::Or::Or(const shared_ptr<Node>& arg0,
-           const shared_ptr<Node>& arg1,
-           const AutoBroadcastSpec& autob)
-    : BinaryElementwiseLogical("Or", arg0, arg1, autob)
+const string op::Or::type_name{"Or"};
+
+op::Or::Or(const Output<Node>& arg0, const Output<Node>& arg1, const AutoBroadcastSpec& autob)
+    : BinaryElementwiseLogical(arg0, arg1, autob)
 {
     constructor_validate_and_infer_types();
 }

src/ngraph/op/or.hpp

@@ -29,6 +29,9 @@ namespace ngraph
         class Or : public util::BinaryElementwiseLogical
         {
         public:
+            NGRAPH_API
+            static const std::string type_name;
+            const std::string& description() const override { return type_name; }
             /// \brief Constructs a logical-or operation.
             ///
             /// \param arg0 Node that produces the first input tensor.<br>
@@ -39,15 +42,14 @@ namespace ngraph
             ///
             /// Output `[d0, ...]`
             ///
-            Or(const std::shared_ptr<Node>& arg0,
-               const std::shared_ptr<Node>& arg1,
+            Or(const Output<Node>& arg0,
+               const Output<Node>& arg1,
                const AutoBroadcastSpec& autob = AutoBroadcastSpec());
 
             virtual std::shared_ptr<Node>
                 copy_with_new_args(const NodeVector& new_args) const override;
 
-        protected:
-            virtual bool is_commutative() override { return true; }
+            virtual bool is_commutative() const override { return true; }
         };
     }
 }