onnx 1.2.1 for BatchNorm and Dropout

Source/CNTKv2LibraryDll/proto/onnx/CNTKToONNX.cpp

@@ -2832,7 +2832,6 @@ void CNTKToONNXHelper::CopyAttributes(const FunctionPtr& src, ONNXIR::Node* node
                 momentum = 1.0f + expm1(-48.0f / normalizationTimeConstant);
 
             node->AddAttribute(attributesMap[L"spatial"], spatial);
-            node->AddAttribute("is_test", (int64_t)1);
             node->AddAttribute(attributesMap[L"epsilon"], epsilon);
             node->AddAttribute("momentum", momentum);
         }
@@ -2877,7 +2876,6 @@ void CNTKToONNXHelper::CopyAttributes(const FunctionPtr& src, ONNXIR::Node* node
         {
             auto dropoutRate = (float)src->Attributes()[L"dropoutRate"].Value<double>();
             node->AddAttribute(attributesMap[L"dropoutRate"], dropoutRate);
-            node->AddAttribute("is_test", (int64_t)1);
         }
         else if ((src->OpName() == L"RandomDistribution") ||
                  (src->OpName() == L"UniformRandom") || (src->OpName() == L"NormalRandom") ||

Source/CNTKv2LibraryDll/proto/onnx/ONNXToCNTK.cpp

@@ -1962,10 +1962,6 @@ FunctionPtr ONNXToCNTKHelper::CreateFunction(const Node *node, const std::vector
     }
     else if (onnxOpName == "BatchNormalization" || onnxOpName == "SpatialBN")
     {
-        auto is_test = GetNamedAttributeAsInt64(node, "is_test", 0);
-        if (is_test == 0)
-            NOT_IMPLEMENTED;
-        // TODO: implement this right once ready.
         const Variable &operand = inputs[0];
         const Variable &scale = inputs[1];
         const Variable &bias = inputs[2];

bindings/python/cntk/tests/onnx_op_test.py

@@ -181,8 +181,6 @@ def test_AveragePool(tmpdir):
 
 #BatchNormalization
 def test_BatchNormalization(tmpdir):
-    pytest.skip('Need to support new ONNX spec.')
-
     dtype = np.float32
 
     sample = [  # 5 samples having 4 classes
@@ -326,7 +324,6 @@ def test_Div(tmpdir):
 
 #Dropout
 def test_Dropout(tmpdir):
-    pytest.skip('Need to support new ONNX spec.')
     data = np.asarray([[10, 20],[30, 40],[50, 60]], dtype=np.float32)
     model = C.dropout(data, 0.5)
     verify_no_input(model, tmpdir, 'Dropout_0')