IVGCVSW-3818 Support Group convolution from Android
Signed-off-by: Teresa Charlin <teresa.charlinreyes@arm.com>
Change-Id: Ie98688175203563ae08019c0c3c876d9c5eae24c
diff --git a/1.2/HalPolicy.cpp b/1.2/HalPolicy.cpp
index d91cb3b..cd3d2a6 100644
--- a/1.2/HalPolicy.cpp
+++ b/1.2/HalPolicy.cpp
@@ -11,6 +11,8 @@
#include <Half.hpp>
#include <TensorUtils.hpp>
+#include <armnn/TypesUtils.hpp>
+
#include <cmath>
namespace armnn_driver
@@ -18,6 +20,8 @@
namespace hal_1_2
{
+using namespace armnn;
+
bool HalPolicy::ConvertOperation(const Operation& operation, const Model& model, ConversionData& data)
{
switch (operation.type)
@@ -48,6 +52,8 @@
return ConvertFloor(operation, model, data);
case V1_2::OperationType::FULLY_CONNECTED:
return ConvertFullyConnected(operation, model, data);
+ case V1_2::OperationType::GROUPED_CONV_2D:
+ return ConvertGroupedConv2d(operation, model, data);
case V1_2::OperationType::L2_NORMALIZATION:
return ConvertL2Normalization(operation, model, data);
case V1_2::OperationType::L2_POOL_2D:
@@ -87,9 +93,9 @@
case V1_2::OperationType::RESHAPE:
return ConvertReshape(operation, model, data);
case V1_2::OperationType::RESIZE_BILINEAR:
- return ConvertResize(operation, model, data, armnn::ResizeMethod::Bilinear);
+ return ConvertResize(operation, model, data, ResizeMethod::Bilinear);
case V1_2::OperationType::RESIZE_NEAREST_NEIGHBOR:
- return ConvertResize(operation, model, data, armnn::ResizeMethod::NearestNeighbor);
+ return ConvertResize(operation, model, data, ResizeMethod::NearestNeighbor);
case V1_2::OperationType::RSQRT:
return ConvertRsqrt(operation, model, data);
case V1_2::OperationType::SQRT:
@@ -133,7 +139,7 @@
bool HalPolicy::ConvertAveragePool2d(const Operation& operation, const Model& model, ConversionData& data)
{
ALOGV("hal_1_2::HalPolicy::ConvertAveragePool2d()");
- return ConvertPooling2d<hal_1_2::HalPolicy>(operation, __func__, armnn::PoolingAlgorithm::Average, model, data);
+ return ConvertPooling2d<hal_1_2::HalPolicy>(operation, __func__, PoolingAlgorithm::Average, model, data);
}
bool HalPolicy::ConvertBatchToSpaceNd(const Operation& operation, const Model& model, ConversionData& data)
@@ -164,16 +170,16 @@
return Fail("%s: Could not read output 0", __func__);
}
- const armnn::TensorInfo& inputInfo = input.GetTensorInfo();
- const armnn::TensorInfo& outputInfo = GetTensorInfoForOperand(*output);
+ const TensorInfo& inputInfo = input.GetTensorInfo();
+ const TensorInfo& outputInfo = GetTensorInfoForOperand(*output);
if (IsDynamicTensor(outputInfo))
{
return Fail("%s: Dynamic output tensors are not supported", __func__);
}
- armnn::Convolution2dDescriptor desc;
- desc.m_DataLayout = armnn::DataLayout::NHWC;
+ Convolution2dDescriptor desc;
+ desc.m_DataLayout = DataLayout::NHWC;
// Determine whether padding is implicit or explicit
bool implicitPadding = operation.inputs.size() == 7 ||
@@ -189,13 +195,13 @@
desc.m_DataLayout = OptionalDataLayout<hal_1_2::HalPolicy>(operation, 10, model, data);
}
- const armnn::PermutationVector OHWIToOIHW = {0, 2, 3, 1};
+ const PermutationVector OHWIToOIHW = {0, 2, 3, 1};
// ArmNN does not currently support non-fixed weights or bias
// The NNAPI filter is always OHWI [depth_out, filter_height, filter_width, depth_in] but ArmNN expects the
// filter's height and width indices to match the input's height and width indices so we permute it to OIHW if
// the DataLayout is NCHW
- const ConstTensorPin weightsPin = (desc.m_DataLayout == armnn::DataLayout::NCHW) ?
+ const ConstTensorPin weightsPin = (desc.m_DataLayout == DataLayout::NCHW) ?
ConvertOperationInputToConstTensorPin<hal_1_2::HalPolicy>(operation, 1, model, data, OHWIToOIHW) :
ConvertOperationInputToConstTensorPin<hal_1_2::HalPolicy>(operation, 1, model, data);
const ConstTensorPin biasPin =
@@ -211,8 +217,8 @@
return Fail("%s: Operation has invalid biases", __func__);
}
- armnn::ConstTensor weights = weightsPin.GetConstTensor();
- armnn::ConstTensor bias = biasPin.GetConstTensor();
+ ConstTensor weights = weightsPin.GetConstTensor();
+ ConstTensor bias = biasPin.GetConstTensor();
SanitizeBiasQuantizationScale(bias.GetInfo(), weights.GetInfo(), inputInfo);
ActivationFn activation;
@@ -262,7 +268,7 @@
}
desc.m_BiasEnabled = true;
- armnn::Optional<armnn::TensorInfo> biases(bias.GetInfo());
+ Optional<TensorInfo> biases(bias.GetInfo());
bool isSupported = false;
FORWARD_LAYER_SUPPORT_FUNC(__func__,
@@ -280,15 +286,15 @@
return false;
}
- armnn::IConnectableLayer* startLayer =
- data.m_Network->AddConvolution2dLayer(desc, weights, armnn::Optional<armnn::ConstTensor>(bias));
+ IConnectableLayer* startLayer =
+ data.m_Network->AddConvolution2dLayer(desc, weights, Optional<ConstTensor>(bias));
if (!startLayer)
{
return Fail("%s: AddConvolution2dLayer failed", __func__);
}
- armnn::IConnectableLayer* endLayer = ProcessActivation(outputInfo, activation, startLayer, data);
+ IConnectableLayer* endLayer = ProcessActivation(outputInfo, activation, startLayer, data);
if (!endLayer)
{
@@ -324,8 +330,8 @@
return Fail("%s: Could not read output 0", __func__);
}
- const armnn::TensorInfo& inputInfo = input.GetTensorInfo();
- const armnn::TensorInfo& outputInfo = GetTensorInfoForOperand(*output);
+ const TensorInfo& inputInfo = input.GetTensorInfo();
+ const TensorInfo& outputInfo = GetTensorInfoForOperand(*output);
if (IsDynamicTensor(outputInfo))
{
@@ -340,8 +346,8 @@
{
return Fail("%s: Operand is invalid", __func__);
}
- armnn::DepthwiseConvolution2dDescriptor desc;
- desc.m_DataLayout = armnn::DataLayout::NHWC;
+ DepthwiseConvolution2dDescriptor desc;
+ desc.m_DataLayout = DataLayout::NHWC;
// Determine whether padding is implicit or explicit
bool implicitPadding = operation.inputs.size() == 8 ||
@@ -358,13 +364,13 @@
unsigned int heightIndex = dataLayoutIndexed.GetHeightIndex();
// Reinterpret weight data as [ H, W, I, M ]
- armnn::TensorShape weightsShape({ weightsOperand->dimensions[1],
+ TensorShape weightsShape({ weightsOperand->dimensions[1],
weightsOperand->dimensions[2],
inputInfo.GetShape()[channelsIndex],
weightsOperand->dimensions[3] / inputInfo.GetShape()[channelsIndex] });
// Swizzle weight data [ H, W, I, M ] -> [ M, I, H, W ]
- const armnn::PermutationVector HWIMToMIHW = { 2U, 3U, 1U, 0U };
+ const PermutationVector HWIMToMIHW = { 2U, 3U, 1U, 0U };
const ConstTensorPin weightsPin =
ConvertOperationInputToConstTensorPin<hal_1_2::HalPolicy>(operation,
@@ -388,8 +394,8 @@
return Fail("%s: Operation has invalid biases", __func__);
}
- armnn::ConstTensor weights = weightsPin.GetConstTensor();
- armnn::ConstTensor bias = biasPin.GetConstTensor();
+ ConstTensor weights = weightsPin.GetConstTensor();
+ ConstTensor bias = biasPin.GetConstTensor();
SanitizeBiasQuantizationScale(bias.GetInfo(), weights.GetInfo(), inputInfo);
ActivationFn activation;
@@ -435,7 +441,7 @@
}
desc.m_BiasEnabled = true;
- armnn::Optional<armnn::TensorInfo> biases(bias.GetInfo());
+ Optional<TensorInfo> biases(bias.GetInfo());
bool isSupported = false;
FORWARD_LAYER_SUPPORT_FUNC(__func__,
@@ -453,15 +459,15 @@
return false;
}
- armnn::IConnectableLayer* startLayer =
- data.m_Network->AddDepthwiseConvolution2dLayer(desc, weights, armnn::Optional<armnn::ConstTensor>(bias));
+ IConnectableLayer* startLayer =
+ data.m_Network->AddDepthwiseConvolution2dLayer(desc, weights, Optional<ConstTensor>(bias));
if (!startLayer)
{
return Fail("%s: AddDepthwiseConvolution2dLayer failed", __func__);
}
- armnn::IConnectableLayer* endLayer = ProcessActivation(outputInfo, activation, startLayer, data);
+ IConnectableLayer* endLayer = ProcessActivation(outputInfo, activation, startLayer, data);
if (!endLayer)
{
return Fail("%s: ProcessActivation failed", __func__);
@@ -501,7 +507,7 @@
return Fail("%s: Operation has invalid output", __func__);
}
- const armnn::TensorInfo& outputInfo = GetTensorInfoForOperand(*output);
+ const TensorInfo& outputInfo = GetTensorInfoForOperand(*output);
if (IsDynamicTensor(outputInfo))
{
return Fail("%s: Dynamic output tensors are not supported", __func__);
@@ -513,7 +519,7 @@
return Fail("%s: failed to get axis input value", __func__);
}
- armnn::TensorShape targetShape;
+ TensorShape targetShape;
try
{
@@ -529,7 +535,7 @@
return Fail("%s: Shape of the output operand does not match the resolved expanded shape", __func__);
}
- armnn::ReshapeDescriptor reshapeDescriptor;
+ ReshapeDescriptor reshapeDescriptor;
reshapeDescriptor.m_TargetShape = targetShape;
bool isSupported = false;
@@ -545,7 +551,7 @@
return false;
}
- armnn::IConnectableLayer* layer = data.m_Network->AddReshapeLayer(reshapeDescriptor);
+ IConnectableLayer* layer = data.m_Network->AddReshapeLayer(reshapeDescriptor);
assert(layer != nullptr);
input.Connect(layer->GetInputSlot(0));
@@ -564,6 +570,322 @@
return ::ConvertFullyConnected<hal_1_2::HalPolicy>(operation, model, data);
}
+bool HalPolicy::ConvertGroupedConv2d(const Operation& operation, const Model& model, ConversionData& data)
+{
+ ALOGV("hal_1_2::HalPolicy::ConvertGroupedConv2d()");
+
+ //
+ // Parse data
+ //
+ LayerInputHandle input = ConvertToLayerInputHandle<hal_1_2::HalPolicy>(operation, 0, model, data);
+ if (!input.IsValid())
+ {
+ return Fail("%s: Operation has invalid inputs", __func__);
+ }
+ const TensorInfo& inputInfo = input.GetTensorInfo();
+
+ const Operand* output = GetOutputOperand<hal_1_2::HalPolicy>(operation, 0, model);
+ if (!output)
+ {
+ return Fail("%s: Could not read output 0", __func__);
+ }
+ const TensorInfo& outputInfo = GetTensorInfoForOperand(*output);
+ if (IsDynamicTensor(outputInfo))
+ {
+ return Fail("%s: Dynamic output tensors are not supported", __func__);
+ }
+
+ // Look ahead to determine data layout
+ DataLayout dataLayout = DataLayout::NHWC;
+ if (operation.inputs.size() == 12)
+ {
+ dataLayout = OptionalDataLayout<hal_1_2::HalPolicy>(operation, 11, model, data);
+ }
+ else
+ {
+ dataLayout = OptionalDataLayout<hal_1_2::HalPolicy>(operation, 8, model, data);
+ }
+
+ // NOTE:
+ // NNAPI weights are always OHWI, i.e. [depth_out, filter_height, filter_width, depth_group],
+ // but Arm NN expects the filter's height and width indices to match the input's height and
+ // width indices so when the DataLayout is NCHW, we need to permute the weights to OIHW
+ const PermutationVector ohwiToOihw = { 0u, 2u, 3u, 1u };
+ const ConstTensorPin weightsPin = (dataLayout == DataLayout::NCHW) ?
+ ConvertOperationInputToConstTensorPin<hal_1_2::HalPolicy>(operation, 1, model, data, ohwiToOihw) :
+ ConvertOperationInputToConstTensorPin<hal_1_2::HalPolicy>(operation, 1, model, data);
+ const ConstTensorPin biasesPin =
+ ConvertOperationInputToConstTensorPin<hal_1_2::HalPolicy>(operation, 2, model, data);
+ if (!weightsPin.IsValid() || !biasesPin.IsValid())
+ {
+ return Fail("%s: Operation has invalid inputs", __func__);
+ }
+
+ ConstTensor weights = weightsPin.GetConstTensor();
+ ConstTensor biases = biasesPin.GetConstTensor();
+ SanitizeBiasQuantizationScale(biases.GetInfo(), weights.GetInfo(), inputInfo);
+
+ const TensorShape& inputShape = inputInfo.GetShape();
+ const TensorShape& outputShape = outputInfo.GetShape();
+ const TensorShape& weightsShape = weights.GetShape();
+ const TensorShape& biasesShape = biases.GetShape();
+
+ armnnUtils::DataLayoutIndexed dataLayoutIndexed(dataLayout);
+ const unsigned int channelsIndex = dataLayoutIndexed.GetChannelsIndex();
+ const unsigned int heightIndex = dataLayoutIndexed.GetHeightIndex();
+ const unsigned int widthIndex = dataLayoutIndexed.GetWidthIndex();
+
+ Convolution2dDescriptor desc;
+ desc.m_DataLayout = dataLayout;
+ desc.m_BiasEnabled = true;
+
+ int numGroups;
+ ActivationFn activation;
+
+ if (operation.inputs.size() == 12)
+ {
+ if (!GetInputScalar<hal_1_2::HalPolicy>(operation, 3, OperandType::INT32, desc.m_PadLeft, model, data) ||
+ !GetInputScalar<hal_1_2::HalPolicy>(operation, 4, OperandType::INT32, desc.m_PadRight, model, data) ||
+ !GetInputScalar<hal_1_2::HalPolicy>(operation, 5, OperandType::INT32, desc.m_PadTop, model, data) ||
+ !GetInputScalar<hal_1_2::HalPolicy>(operation, 6, OperandType::INT32, desc.m_PadBottom, model, data) ||
+ !GetInputScalar<hal_1_2::HalPolicy>(operation, 7, OperandType::INT32, desc.m_StrideX, model, data) ||
+ !GetInputScalar<hal_1_2::HalPolicy>(operation, 8, OperandType::INT32, desc.m_StrideY, model, data) ||
+ !GetInputScalar<hal_1_2::HalPolicy>(operation, 9, OperandType::INT32, numGroups, model, data) ||
+ !GetInputActivationFunction<hal_1_2::HalPolicy>(operation, 10, activation, model, data))
+ {
+ return Fail("%s: Operation has invalid inputs (explicit padding)", __func__);
+ }
+
+ }
+ else if (operation.inputs.size() == 9)
+ {
+ android::nn::PaddingScheme paddingScheme;
+ if (!GetInputPaddingScheme<hal_1_2::HalPolicy>(operation, 3, paddingScheme, model, data) ||
+ !GetInputScalar<hal_1_2::HalPolicy>(operation, 4, OperandType::INT32, desc.m_StrideX, model, data) ||
+ !GetInputScalar<hal_1_2::HalPolicy>(operation, 5, OperandType::INT32, desc.m_StrideY, model, data) ||
+ !GetInputScalar<hal_1_2::HalPolicy>(operation, 6, OperandType::INT32, numGroups, model, data) ||
+ !GetInputActivationFunction<hal_1_2::HalPolicy>(operation, 7, activation, model, data))
+ {
+ return Fail("%s: Operation has invalid inputs (implicit padding)", __func__);
+ }
+
+ const uint32_t inputX = inputInfo.GetShape()[widthIndex];
+ const uint32_t inputY = inputInfo.GetShape()[heightIndex];
+
+ const uint32_t kernelX = weightsShape[widthIndex];
+ const uint32_t kernelY = weightsShape[heightIndex];
+
+ CalcPadding(inputX, kernelX, desc.m_StrideX, desc.m_PadLeft, desc.m_PadRight, paddingScheme);
+ CalcPadding(inputY, kernelY, desc.m_StrideY, desc.m_PadTop, desc.m_PadBottom, paddingScheme);
+ }
+ else
+ {
+ return Fail("%s: Unsupported number of operation inputs", __func__);
+ }
+
+ const unsigned int outputChannels = outputShape[channelsIndex];
+
+ const unsigned int channelsPerGroup = weightsShape[channelsIndex];
+ const unsigned int channelMultiplier = outputChannels / numGroups;
+
+ //
+ // Validate all relevant inputs
+ //
+ if (numGroups <= 0)
+ {
+ return Fail("%s: Number of groups must be greater than 0. Got: %d", __func__, numGroups);
+ }
+
+ if (outputChannels % numGroups != 0u)
+ {
+ return Fail("%s: Output channels must be divisible by the number of groups", __func__);
+ }
+
+ //
+ // Set up Splitter layer
+ //
+ unsigned int splitterDimSizes[4] = { inputShape[0], inputShape[1], inputShape[2], inputShape[3] };
+ splitterDimSizes[channelsIndex] /= numGroups; // split in depth
+
+ TensorInfo splitterOutputInfo(4,
+ splitterDimSizes,
+ inputInfo.GetDataType(),
+ inputInfo.GetQuantizationScale(),
+ inputInfo.GetQuantizationOffset());
+
+ std::vector<std::reference_wrapper<TensorInfo>> splitterOutputInfos(numGroups, std::ref(splitterOutputInfo));
+
+ ViewsDescriptor splitterDesc(numGroups);
+ for (unsigned int group = 0u; group < numGroups; ++group)
+ {
+ splitterDesc.SetViewOriginCoord(group, channelsIndex, splitterDimSizes[channelsIndex] * group);
+ for (unsigned int dimIdx = 0u; dimIdx < 4u; dimIdx++)
+ {
+ splitterDesc.SetViewSize(group, dimIdx, splitterDimSizes[dimIdx]);
+ }
+ }
+
+ bool isSupported = false;
+ FORWARD_LAYER_SUPPORT_FUNC(__func__,
+ IsSplitterSupported,
+ data.m_Backends,
+ isSupported,
+ inputInfo,
+ splitterOutputInfos,
+ splitterDesc);
+ if (!isSupported)
+ {
+ return false;
+ }
+
+ IConnectableLayer* splitterLayer = data.m_Network->AddSplitterLayer(splitterDesc);
+ if (!splitterLayer)
+ {
+ return Fail("%s: Failed to add SplitterLayer", __func__);
+ }
+
+ input.Connect(splitterLayer->GetInputSlot(0));
+ for (unsigned int group = 0u; group < splitterLayer->GetNumOutputSlots(); ++group)
+ {
+ splitterLayer->GetOutputSlot(group).SetTensorInfo(splitterOutputInfo);
+ }
+
+ //
+ // Set up Convolution2d layers for each group
+ //
+ TensorShape groupInputShape(inputShape);
+ groupInputShape[channelsIndex] = channelsPerGroup;
+
+ TensorShape groupOutputShape(outputShape);
+ groupOutputShape[channelsIndex] = 1;
+
+ TensorShape groupWeightsShape(weightsShape);
+ groupWeightsShape[0] /= channelMultiplier * numGroups;
+
+ TensorShape groupBiasesShape({ 1 });
+
+ const TensorInfo groupInputInfo (groupInputShape,
+ inputInfo.GetDataType(),
+ inputInfo.GetQuantizationScale(),
+ inputInfo.GetQuantizationOffset());
+ const TensorInfo groupWeightsInfo(groupWeightsShape,
+ weights.GetInfo().GetDataType(),
+ weights.GetInfo().GetQuantizationScale(),
+ weights.GetInfo().GetQuantizationOffset());
+ const TensorInfo groupBiasesInfo (groupBiasesShape,
+ biases.GetInfo().GetDataType(),
+ biases.GetInfo().GetQuantizationScale(),
+ biases.GetInfo().GetQuantizationOffset());
+ const TensorInfo groupOutputInfo (groupOutputShape,
+ outputInfo.GetDataType(),
+ outputInfo.GetQuantizationScale(),
+ outputInfo.GetQuantizationOffset());
+
+ const unsigned int weightsDataTypeSize = GetDataTypeSize(groupWeightsInfo.GetDataType());
+ const unsigned int biasesDataTypeSize = GetDataTypeSize(groupBiasesInfo.GetDataType());
+
+ std::vector<IConnectableLayer*> convLayers(numGroups*channelMultiplier, nullptr);
+ for (unsigned int group = 0u; group < numGroups; ++group)
+ {
+ for (unsigned int m = 0u; m < channelMultiplier; ++m)
+ {
+ auto index = group * channelMultiplier + m;
+
+ const unsigned int weightsDataOffset = groupWeightsShape.GetNumElements() * index * weightsDataTypeSize;
+ const unsigned int biasesDataOffset = groupBiasesShape.GetNumElements() * index * biasesDataTypeSize;
+
+ // Extract weights and biases data for current group convolution
+ ConstTensor groupWeights(groupWeightsInfo,
+ static_cast<const void *>(reinterpret_cast<const char *>(weights.GetMemoryArea()) +
+ weightsDataOffset));
+ ConstTensor groupBiases(groupBiasesInfo,
+ static_cast<const void *>(reinterpret_cast<const char *>(biases.GetMemoryArea()) +
+ biasesDataOffset));
+
+ isSupported = false;
+ FORWARD_LAYER_SUPPORT_FUNC(__func__,
+ IsConvolution2dSupported,
+ data.m_Backends,
+ isSupported,
+ groupInputInfo,
+ groupOutputInfo,
+ desc,
+ groupWeightsInfo,
+ Optional<TensorInfo>(groupBiasesInfo));
+ if (!isSupported)
+ {
+ return false;
+ }
+
+ IConnectableLayer *convLayer =
+ data.m_Network->AddConvolution2dLayer(desc, groupWeights, Optional<ConstTensor>(groupBiases));
+ if (!convLayer)
+ {
+ return Fail("%s: AddConvolution2dLayer failed", __func__);
+ }
+
+ splitterLayer->GetOutputSlot(group).Connect(convLayer->GetInputSlot(0));
+ convLayer->GetOutputSlot(0).SetTensorInfo(groupOutputInfo);
+
+ convLayers[index] = convLayer;
+ }
+ }
+
+ //
+ // Set up Concat layer
+ //
+ ConcatDescriptor concatDescriptor(outputInfo.GetShape()[channelsIndex]);
+ for (unsigned int group = 0u; group < numGroups; ++group)
+ {
+ for (unsigned int m = 0u; m < channelMultiplier; ++m)
+ {
+ auto index = group * channelMultiplier + m;
+ concatDescriptor.SetViewOriginCoord(index, channelsIndex, index);
+ concatDescriptor.SetConcatAxis(channelsIndex);
+ }
+ }
+
+ isSupported = false;
+ FORWARD_LAYER_SUPPORT_FUNC(__func__,
+ IsConcatSupported,
+ data.m_Backends,
+ isSupported,
+ std::vector<const TensorInfo*>(numGroups * channelMultiplier, &groupOutputInfo),
+ outputInfo,
+ concatDescriptor);
+ if (!isSupported)
+ {
+ return false;
+ }
+
+ IConnectableLayer* concatLayer = data.m_Network->AddConcatLayer(concatDescriptor);
+ if (!concatLayer)
+ {
+ return Fail("%s: AddConcatLayer failed", __func__);
+ }
+
+ for (unsigned int group = 0u; group < numGroups; ++group)
+ {
+ for (unsigned int m = 0u; m < channelMultiplier; ++m)
+ {
+ auto index = group * channelMultiplier + m;
+ convLayers[index]->GetOutputSlot(0).Connect(concatLayer->GetInputSlot(index));
+ }
+ }
+ concatLayer->GetOutputSlot(0).SetTensorInfo(outputInfo);
+
+ //
+ // Set up Activation layer (if it is set)
+ //
+ IConnectableLayer* endLayer = ProcessActivation(outputInfo, activation, concatLayer, data);
+ if (!endLayer)
+ {
+ return Fail("%s: ProcessActivation failed", __func__);
+ }
+
+ return SetupAndTrackLayerOutputSlot<HalPolicy>(operation, 0, *endLayer, model, data);
+}
+
bool HalPolicy::ConvertL2Normalization(const Operation& operation, const Model& model, ConversionData& data)
{
ALOGV("hal_1_2::HalPolicy::ConvertL2Normalization()");
@@ -573,7 +895,7 @@
bool HalPolicy::ConvertL2Pool2d(const Operation& operation, const Model& model, ConversionData& data)
{
ALOGV("hal_1_2::HalPolicy::ConvertL2Pool2d()");
- return ConvertPooling2d<hal_1_2::HalPolicy>(operation, __func__, armnn::PoolingAlgorithm::L2, model, data);
+ return ConvertPooling2d<hal_1_2::HalPolicy>(operation, __func__, PoolingAlgorithm::L2, model, data);
}
bool HalPolicy::ConvertLocalResponseNormalization(const Operation& operation,
@@ -593,7 +915,7 @@
bool HalPolicy::ConvertMaxPool2d(const Operation& operation, const Model& model, ConversionData& data)
{
ALOGV("hal_1_2::HalPolicy::ConvertMaxPool2d()");
- return ConvertPooling2d<hal_1_2::HalPolicy>(operation, __func__, armnn::PoolingAlgorithm::Max, model, data);
+ return ConvertPooling2d<hal_1_2::HalPolicy>(operation, __func__, PoolingAlgorithm::Max, model, data);
}
bool HalPolicy::ConvertMaximum(const Operation& operation, const Model& model, ConversionData& data)
@@ -614,7 +936,7 @@
return Fail("%s: Could not read output", __func__);
}
- const armnn::TensorInfo& outInfo = GetTensorInfoForOperand(*outputOperand);
+ const TensorInfo& outInfo = GetTensorInfoForOperand(*outputOperand);
if (IsDynamicTensor(outInfo))
{
return Fail("%s: Dynamic output tensors are not supported", __func__);
@@ -634,7 +956,7 @@
return false;
}
- armnn::IConnectableLayer* layer = data.m_Network->AddMaximumLayer();
+ IConnectableLayer* layer = data.m_Network->AddMaximumLayer();
assert(layer != nullptr);
bool isReshapeSupported = BroadcastTensor(input0, input1, layer, data);
if (!isReshapeSupported)
@@ -669,7 +991,7 @@
return Fail("%s: Could not read output 0", __func__);
}
- const armnn::TensorInfo& outputInfo = GetTensorInfoForOperand(*output);
+ const TensorInfo& outputInfo = GetTensorInfoForOperand(*output);
if (IsDynamicTensor(outputInfo))
{
return Fail("%s: Dynamic output tensors are not supported", __func__);
@@ -689,7 +1011,7 @@
return false;
}
- armnn::IConnectableLayer* const layer = data.m_Network->AddMinimumLayer();
+ IConnectableLayer* const layer = data.m_Network->AddMinimumLayer();
assert(layer != nullptr);
bool isReshapeSupported = BroadcastTensor(input0, input1, layer, data);
if (!isReshapeSupported)
@@ -728,16 +1050,16 @@
return Fail("%s: Could not read output", __func__);
}
- const armnn::TensorInfo& inputInfo = input.GetTensorInfo();
+ const TensorInfo& inputInfo = input.GetTensorInfo();
unsigned int rank = inputInfo.GetNumDimensions();
- armnn::PadDescriptor descriptor;
+ PadDescriptor descriptor;
if (!ConvertPaddings<hal_1_2::HalPolicy>(operation, model, data, rank, descriptor))
{
return Fail("%s: Could not convert paddings", __func__);
}
- const armnn::TensorInfo& outputInfo = GetTensorInfoForOperand(*output);
+ const TensorInfo& outputInfo = GetTensorInfoForOperand(*output);
if (IsDynamicTensor(outputInfo))
{
return Fail("%s: Dynamic output tensors are not supported", __func__);
@@ -756,7 +1078,7 @@
// Read value to use for padding
if (operandType0 == OperandType::TENSOR_FLOAT16 && operandType2 == OperandType::FLOAT16)
{
- armnn::Half f16PadValue;
+ Half f16PadValue;
if (!GetInputScalar<hal_1_2::HalPolicy>(operation, 2, operandType2, f16PadValue, model, data))
{
return Fail("%s: Could not read input 2 (FLOAT16)", __func__);
@@ -798,7 +1120,7 @@
return false;
}
- armnn::IConnectableLayer* const layer = data.m_Network->AddPadLayer(descriptor);
+ IConnectableLayer* const layer = data.m_Network->AddPadLayer(descriptor);
assert(layer != nullptr);
input.Connect(layer->GetInputSlot(0));
layer->GetOutputSlot(0).SetTensorInfo(outputInfo);
@@ -825,9 +1147,9 @@
return Fail("%s: Could not read output", __func__);
}
- const armnn::TensorInfo& inputInfo = input.GetTensorInfo();
- const armnn::TensorInfo& alphaInfo = alpha.GetTensorInfo();
- const armnn::TensorInfo& outputInfo = GetTensorInfoForOperand(*output);
+ const TensorInfo& inputInfo = input.GetTensorInfo();
+ const TensorInfo& alphaInfo = alpha.GetTensorInfo();
+ const TensorInfo& outputInfo = GetTensorInfoForOperand(*output);
if (IsDynamicTensor(outputInfo))
{
@@ -847,7 +1169,7 @@
return false;
}
- armnn::IConnectableLayer* const layer = data.m_Network->AddPreluLayer();
+ IConnectableLayer* const layer = data.m_Network->AddPreluLayer();
if (!layer)
{
@@ -879,7 +1201,7 @@
return Fail("%s: Operation has invalid outputs", __func__);
}
- const armnn::TensorInfo& outputInfo = GetTensorInfoForOperand(*outputOperand);
+ const TensorInfo& outputInfo = GetTensorInfoForOperand(*outputOperand);
if (IsDynamicTensor(outputInfo))
{
return Fail("%s: Dynamic output tensors are not supported", __func__);
@@ -897,7 +1219,7 @@
return false;
}
- armnn::IConnectableLayer* const layer = data.m_Network->AddQuantizeLayer();
+ IConnectableLayer* const layer = data.m_Network->AddQuantizeLayer();
assert(layer != nullptr);
input.Connect(layer->GetInputSlot(0));
@@ -1043,13 +1365,13 @@
}
// Inputs
- const armnn::TensorInfo& inputInfo = input.GetTensorInfo();
- const armnn::TensorInfo& previousCellStateInInfo = previousCellStateIn.GetTensorInfo();
- const armnn::TensorInfo& previousOutputInInfo = previousOutputIn.GetTensorInfo();
+ const TensorInfo& inputInfo = input.GetTensorInfo();
+ const TensorInfo& previousCellStateInInfo = previousCellStateIn.GetTensorInfo();
+ const TensorInfo& previousOutputInInfo = previousOutputIn.GetTensorInfo();
// Outputs
- const armnn::TensorInfo& cellStateOutInfo = GetTensorInfoForOperand(*cellStateOut);
- const armnn::TensorInfo& outputInfo = GetTensorInfoForOperand(*output);
+ const TensorInfo& cellStateOutInfo = GetTensorInfoForOperand(*cellStateOut);
+ const TensorInfo& outputInfo = GetTensorInfoForOperand(*output);
// Dynamic tensors currently not supported
if (IsDynamicTensor(cellStateOutInfo) || IsDynamicTensor(outputInfo))
@@ -1057,7 +1379,7 @@
return Fail("%s: Dynamic output tensors are not supported", __func__);
}
- armnn::QuantizedLstmInputParams params;
+ QuantizedLstmInputParams params;
params.m_InputToInputWeights = inputToInputWeightsPin.GetConstTensorPtr();
params.m_InputToForgetWeights = inputToForgetWeightsPin.GetConstTensorPtr();
@@ -1072,7 +1394,7 @@
params.m_CellBias = cellBiasPin.GetConstTensorPtr();
params.m_OutputGateBias = outputGateBiasPin.GetConstTensorPtr();
- armnn::QuantizedLstmInputParamsInfo paramsInfo;
+ QuantizedLstmInputParamsInfo paramsInfo;
paramsInfo.m_InputToInputWeights = &(params.m_InputToInputWeights->GetInfo());
paramsInfo.m_InputToForgetWeights = &(params.m_InputToForgetWeights->GetInfo());
paramsInfo.m_InputToCellWeights = &(params.m_InputToCellWeights->GetInfo());
@@ -1103,7 +1425,7 @@
return false;
}
- armnn::IConnectableLayer* const layer = data.m_Network->AddQuantizedLstmLayer(params, "QuantizedLstm");
+ IConnectableLayer* const layer = data.m_Network->AddQuantizedLstmLayer(params, "QuantizedLstm");
input.Connect(layer->GetInputSlot(0));
previousCellStateIn.Connect(layer->GetInputSlot(1));
previousOutputIn.Connect(layer->GetInputSlot(2));
@@ -1139,7 +1461,7 @@
bool HalPolicy::ConvertResize(const Operation& operation,
const Model& model,
ConversionData& data,
- armnn::ResizeMethod resizeMethod)
+ ResizeMethod resizeMethod)
{
ALOGV("hal_1_2::HalPolicy::ConvertResize()");
@@ -1155,15 +1477,15 @@
return Fail("%s: Could not read output 0", __func__);
}
- const armnn::TensorInfo& inputInfo = input.GetTensorInfo();
- const armnn::TensorInfo& outputInfo = GetTensorInfoForOperand(*output);
+ const TensorInfo& inputInfo = input.GetTensorInfo();
+ const TensorInfo& outputInfo = GetTensorInfoForOperand(*output);
if (IsDynamicTensor(outputInfo))
{
return Fail("%s: Dynamic output tensors are not supported", __func__);
}
- armnn::ResizeDescriptor descriptor;
+ ResizeDescriptor descriptor;
descriptor.m_Method = resizeMethod;
descriptor.m_DataLayout = OptionalDataLayout<hal_1_2::HalPolicy>(operation, 3, model, data);
@@ -1214,7 +1536,7 @@
return Fail("%s: Operation has invalid inputs for resizing by scale", __func__);
}
- const armnn::TensorShape& inputShape = inputInfo.GetShape();
+ const TensorShape& inputShape = inputInfo.GetShape();
armnnUtils::DataLayoutIndexed dataLayoutIndexed(descriptor.m_DataLayout);
float width = inputShape[dataLayoutIndexed.GetWidthIndex()];
@@ -1243,7 +1565,7 @@
return false;
}
- armnn::IConnectableLayer* layer = data.m_Network->AddResizeLayer(descriptor);
+ IConnectableLayer* layer = data.m_Network->AddResizeLayer(descriptor);
assert(layer != nullptr);
@@ -1269,7 +1591,7 @@
return Fail("%s: Could not read output 0", __func__);
}
- const armnn::TensorInfo& outputInfo = GetTensorInfoForOperand(*output);
+ const TensorInfo& outputInfo = GetTensorInfoForOperand(*output);
if (IsDynamicTensor(outputInfo))
{
return Fail("%s: Dynamic output tensors are not supported", __func__);
@@ -1288,7 +1610,7 @@
return false;
}
- armnn::IConnectableLayer* const layer = data.m_Network->AddRsqrtLayer();
+ IConnectableLayer* const layer = data.m_Network->AddRsqrtLayer();
assert(layer != nullptr);
input.Connect(layer->GetInputSlot(0));
@@ -1311,7 +1633,7 @@
return Fail("%s: Operation has invalid inputs", __func__);
}
- const armnn::TensorInfo& inputInfo = input.GetTensorInfo();
+ const TensorInfo& inputInfo = input.GetTensorInfo();
unsigned int rank = inputInfo.GetNumDimensions();
if (rank != 4)
{
@@ -1324,13 +1646,13 @@
return Fail("%s: Could not read output 0", __func__);
}
- const armnn::TensorInfo& outputInfo = GetTensorInfoForOperand(*output);
+ const TensorInfo& outputInfo = GetTensorInfoForOperand(*output);
if (IsDynamicTensor(outputInfo))
{
return Fail("%s: Dynamic output tensors are not supported", __func__);
}
- armnn::SpaceToDepthDescriptor desc;
+ SpaceToDepthDescriptor desc;
GetInputScalar<hal_1_2::HalPolicy>(operation, 1, OperandType::INT32, desc.m_BlockSize, model, data);
@@ -1354,7 +1676,7 @@
return false;
}
- armnn::IConnectableLayer* const layer = data.m_Network->AddSpaceToDepthLayer(desc);
+ IConnectableLayer* const layer = data.m_Network->AddSpaceToDepthLayer(desc);
assert(layer != nullptr);
input.Connect(layer->GetInputSlot(0));
@@ -1377,13 +1699,13 @@
return Fail("%s: Operation has no outputs", __func__);
}
- const armnn::TensorInfo& outputInfo = GetTensorInfoForOperand(*outputOperand);
+ const TensorInfo& outputInfo = GetTensorInfoForOperand(*outputOperand);
if (IsDynamicTensor(outputInfo))
{
return Fail("%s: Dynamic output tensors are not supported", __func__);
}
- armnn::SoftmaxDescriptor desc;
+ SoftmaxDescriptor desc;
if (!GetInputFloat32<hal_1_2::HalPolicy>(operation, 1, desc.m_Beta, model, data))
{
return Fail("%s: Operation has invalid inputs", __func__);
@@ -1419,7 +1741,7 @@
return false;
}
- armnn::IConnectableLayer* layer = data.m_Network->AddSoftmaxLayer(desc);
+ IConnectableLayer* layer = data.m_Network->AddSoftmaxLayer(desc);
assert(layer != nullptr);
input.Connect(layer->GetInputSlot(0));
@@ -1697,7 +2019,7 @@
}
// set the params structure for the AddLstmLayer call
- armnn::LstmInputParams params;
+ LstmInputParams params;
params.m_InputToInputWeights = inputToInputWeightsPin.GetConstTensorPtr();
params.m_InputToForgetWeights = inputToForgetWeightsPin.GetConstTensorPtr();
params.m_InputToCellWeights = inputToCellWeightsPin.GetConstTensorPtr();
@@ -1721,7 +2043,7 @@
params.m_OutputLayerNormWeights = outputLayerNormWeightsPin.GetConstTensorPtr();
// set the layer descriptor
- armnn::LstmDescriptor desc;
+ LstmDescriptor desc;
desc.m_ActivationFunc = activation;
desc.m_ClippingThresCell = cellClip;
desc.m_ClippingThresProj = projClip;
@@ -1772,15 +2094,15 @@
// Check if the layer is supported
// Inputs
- const armnn::TensorInfo& inputInfo = input.GetTensorInfo();
- const armnn::TensorInfo& outputStateInInfo = outputStateIn.GetTensorInfo();
- const armnn::TensorInfo& cellStateInInfo = cellStateIn.GetTensorInfo();
+ const TensorInfo& inputInfo = input.GetTensorInfo();
+ const TensorInfo& outputStateInInfo = outputStateIn.GetTensorInfo();
+ const TensorInfo& cellStateInInfo = cellStateIn.GetTensorInfo();
// Outputs
- const armnn::TensorInfo& scratchBufferInfo = GetTensorInfoForOperand(*scratchBuffer);
- const armnn::TensorInfo& outputStateOutInfo = GetTensorInfoForOperand(*outputStateOut);
- const armnn::TensorInfo& cellStateOutInfo = GetTensorInfoForOperand(*cellStateOut);
- const armnn::TensorInfo& outputInfo = GetTensorInfoForOperand(*output);
+ const TensorInfo& scratchBufferInfo = GetTensorInfoForOperand(*scratchBuffer);
+ const TensorInfo& outputStateOutInfo = GetTensorInfoForOperand(*outputStateOut);
+ const TensorInfo& cellStateOutInfo = GetTensorInfoForOperand(*cellStateOut);
+ const TensorInfo& outputInfo = GetTensorInfoForOperand(*output);
if (IsDynamicTensor(scratchBufferInfo) ||
IsDynamicTensor(outputStateOutInfo) ||
@@ -1791,7 +2113,7 @@
}
// Basic parameters
- armnn::LstmInputParamsInfo paramsInfo;
+ LstmInputParamsInfo paramsInfo;
paramsInfo.m_InputToForgetWeights = &(params.m_InputToForgetWeights->GetInfo());
paramsInfo.m_InputToCellWeights = &(params.m_InputToCellWeights->GetInfo());
paramsInfo.m_InputToOutputWeights = &(params.m_InputToOutputWeights->GetInfo());
@@ -1860,7 +2182,7 @@
}
// Add the layer
- armnn::IConnectableLayer* layer = data.m_Network->AddLstmLayer(desc, params, "Lstm");
+ IConnectableLayer* layer = data.m_Network->AddLstmLayer(desc, params, "Lstm");
input.Connect(layer->GetInputSlot(0));
outputStateIn.Connect(layer->GetInputSlot(1));
@@ -1875,8 +2197,8 @@
bool HalPolicy::ConvertSqrt(const Operation& operation, const Model& model, ConversionData& data)
{
ALOGV("hal_1_2::HalPolicy::ConvertSqrt()");
- armnn::ActivationDescriptor desc;
- desc.m_Function = armnn::ActivationFunction::Sqrt;
+ ActivationDescriptor desc;
+ desc.m_Function = ActivationFunction::Sqrt;
return ::ConvertToActivation<hal_1_2::HalPolicy>(operation, __func__, desc, model, data);
}
@@ -1915,8 +2237,8 @@
return Fail("%s: Could not read output 0", __func__);
}
- const armnn::TensorInfo& inputInfo = input.GetTensorInfo();
- const armnn::TensorInfo& outputInfo = GetTensorInfoForOperand(*output);
+ const TensorInfo& inputInfo = input.GetTensorInfo();
+ const TensorInfo& outputInfo = GetTensorInfoForOperand(*output);
if (IsDynamicTensor(outputInfo))
{
return Fail("%s: Dynamic output tensors are not supported", __func__);
@@ -1930,8 +2252,8 @@
{
return Fail("%s: Operand is invalid", __func__);
}
- armnn::TransposeConvolution2dDescriptor desc;
- desc.m_DataLayout = armnn::DataLayout::NHWC;
+ TransposeConvolution2dDescriptor desc;
+ desc.m_DataLayout = DataLayout::NHWC;
// Determine whether padding is implicit or explicit
bool implicitPadding = operation.inputs.size() == 9;
@@ -1949,11 +2271,11 @@
unsigned int widthIndex = dataLayoutIndexed.GetWidthIndex();
unsigned int heightIndex = dataLayoutIndexed.GetHeightIndex();
- const armnn::PermutationVector OHWIToOIHW = {0, 2, 3, 1};
+ const PermutationVector OHWIToOIHW = {0, 2, 3, 1};
// The shape of the weight is [depth_out, filter_height, filter_width, depth_in].
// We have to permute it to OIHW if the data layout is NCHW.
- const ConstTensorPin weightsPin = (desc.m_DataLayout == armnn::DataLayout::NCHW) ?
+ const ConstTensorPin weightsPin = (desc.m_DataLayout == DataLayout::NCHW) ?
ConvertOperationInputToConstTensorPin<hal_1_2::HalPolicy>(operation, 1, model, data, OHWIToOIHW) :
ConvertOperationInputToConstTensorPin<hal_1_2::HalPolicy>(operation, 1, model, data);
@@ -1971,8 +2293,8 @@
return Fail("%s: Operation has invalid biases", __func__);
}
- armnn::ConstTensor weights = weightsPin.GetConstTensor();
- armnn::ConstTensor bias = biasPin.GetConstTensor();
+ ConstTensor weights = weightsPin.GetConstTensor();
+ ConstTensor bias = biasPin.GetConstTensor();
SanitizeBiasQuantizationScale(bias.GetInfo(), weights.GetInfo(), inputInfo);
ActivationFn activation;
@@ -2037,7 +2359,7 @@
}
desc.m_BiasEnabled = true;
- armnn::Optional<armnn::TensorInfo> biases(bias.GetInfo());
+ Optional<TensorInfo> biases(bias.GetInfo());
bool isSupported = false;
FORWARD_LAYER_SUPPORT_FUNC(__func__,
@@ -2054,14 +2376,14 @@
return false;
}
- armnn::IConnectableLayer* startLayer =
- data.m_Network->AddTransposeConvolution2dLayer(desc, weights, armnn::Optional<armnn::ConstTensor>(bias));
+ IConnectableLayer* startLayer =
+ data.m_Network->AddTransposeConvolution2dLayer(desc, weights, Optional<ConstTensor>(bias));
if (!startLayer)
{
return Fail("%s: AddTransposeConvolution2dLayer failed", __func__);
}
- armnn::IConnectableLayer* endLayer = ProcessActivation(outputInfo, activation, startLayer, data);
+ IConnectableLayer* endLayer = ProcessActivation(outputInfo, activation, startLayer, data);
if (!endLayer)
{
return Fail("%s: ProcessActivation failed", __func__);
diff --git a/1.2/HalPolicy.hpp b/1.2/HalPolicy.hpp
index 82de44c..52f4a42 100644
--- a/1.2/HalPolicy.hpp
+++ b/1.2/HalPolicy.hpp
@@ -57,6 +57,8 @@
static bool ConvertFullyConnected(const Operation& operation, const Model& model, ConversionData& data);
+ static bool ConvertGroupedConv2d(const Operation& operation, const Model& model, ConversionData& data);
+
static bool ConvertL2Normalization(const Operation& operation, const Model& model, ConversionData& data);
static bool ConvertL2Pool2d(const Operation& operation, const Model& model, ConversionData& data);
diff --git a/ConversionUtils.hpp b/ConversionUtils.hpp
index 2e4cadd..de25580 100644
--- a/ConversionUtils.hpp
+++ b/ConversionUtils.hpp
@@ -125,43 +125,46 @@
// Convenience macro to call an Is*Supported function and log caller name together with reason for lack of support.
// Called as: FORWARD_LAYER_SUPPORT_FUNC(__func__, Is*Supported, backends, a, b, c, d, e)
#define FORWARD_LAYER_SUPPORT_FUNC(funcName, func, backends, supported, ...) \
- std::string reasonIfUnsupported; \
- try { \
- for (auto&& backendId : backends) \
+try \
+{ \
+ for (auto&& backendId : backends) \
+ { \
+ auto layerSupportObject = armnn::GetILayerSupportByBackendId(backendId); \
+ if (layerSupportObject) \
{ \
- auto layerSupportObject = armnn::GetILayerSupportByBackendId(backendId); \
- if (layerSupportObject) \
+ std::string reasonIfUnsupported; \
+ supported = \
+ layerSupportObject->func(__VA_ARGS__, armnn::Optional<std::string&>(reasonIfUnsupported)); \
+ if (supported) \
{ \
- supported = \
- layerSupportObject->func(__VA_ARGS__, armnn::Optional<std::string&>(reasonIfUnsupported)); \
- if (supported) \
- { \
- break; \
- } \
- else \
- { \
- if (reasonIfUnsupported.size() > 0) \
- { \
- ALOGD("%s: not supported by armnn: %s", funcName, reasonIfUnsupported.c_str()); \
- } \
- else \
- { \
- ALOGD("%s: not supported by armnn", funcName); \
- } \
- } \
+ break; \
} \
else \
{ \
- ALOGD("%s: backend not registered: %s", funcName, backendId.Get().c_str()); \
+ if (reasonIfUnsupported.size() > 0) \
+ { \
+ ALOGD("%s: not supported by armnn: %s", funcName, reasonIfUnsupported.c_str()); \
+ } \
+ else \
+ { \
+ ALOGD("%s: not supported by armnn", funcName); \
+ } \
} \
} \
- if (!supported) \
+ else \
{ \
- ALOGD("%s: not supported by any specified backend", funcName); \
+ ALOGD("%s: backend not registered: %s", funcName, backendId.Get().c_str()); \
} \
- } catch (const armnn::InvalidArgumentException &e) { \
- throw armnn::InvalidArgumentException(e, "Failed to check layer support", CHECK_LOCATION()); \
- }
+ } \
+ if (!supported) \
+ { \
+ ALOGD("%s: not supported by any specified backend", funcName); \
+ } \
+} \
+catch (const armnn::InvalidArgumentException &e) \
+{ \
+ throw armnn::InvalidArgumentException(e, "Failed to check layer support", CHECK_LOCATION()); \
+}
template<typename Operand>
armnn::TensorShape GetTensorShapeForOperand(const Operand& operand)
diff --git a/NnapiSupport.txt b/NnapiSupport.txt
index 3dea13e..be36514 100644
--- a/NnapiSupport.txt
+++ b/NnapiSupport.txt
@@ -25,6 +25,7 @@
EXPAND_DIMS (FLOAT32,QUANT8_ASYMM)
FLOOR (FLOAT32)
FULLY_CONNECTED (FLOAT32,QUANT8_ASYMM)
+GROUPED_CONV_2D (FLOAT32,QUANT8_ASYMM)
L2_NORMALIZATION (FLOAT32)
L2_POOL_2D (FLOAT32,QUANT8_ASYMM)
LOCAL_RESPONSE_NORMALIZATION (FLOAT32)
