COMPMID-417 - Added DirectConvolution 5x5 in AlexNet system test
Change-Id: I76622fbce993df5d16eba10d31813bc196ce110a
Reviewed-on: http://mpd-gerrit.cambridge.arm.com/84772
Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
Tested-by: Kaizen <jeremy.johnson+kaizengerrit@arm.com>
diff --git a/tests/networks_new/AlexNetNetwork.h b/tests/networks_new/AlexNetNetwork.h
index 28bcfa1..39c69da 100644
--- a/tests/networks_new/AlexNetNetwork.h
+++ b/tests/networks_new/AlexNetNetwork.h
@@ -98,27 +98,26 @@
const unsigned int data_type_size = 16 / arm_compute::data_size_from_type(_data_type);
// Create tensor for the reshaped weights
- auto w21_tensor = std::unique_ptr<TensorType>(new TensorType());
- auto w22_tensor = std::unique_ptr<TensorType>(new TensorType());
-
w[0].allocator()->init(TensorInfo(TensorShape(366U * data_type_size, 96U / data_type_size), 1, _data_type, _fixed_point_position));
- w21_tensor->allocator()->init(TensorInfo(TensorShape(1248U * data_type_size, 128U / data_type_size), 1, _data_type, _fixed_point_position));
- w22_tensor->allocator()->init(TensorInfo(TensorShape(1248U * data_type_size, 128U / data_type_size), 1, _data_type, _fixed_point_position));
- w21 = std::move(w21_tensor);
- w22 = std::move(w22_tensor);
// Configure the direct convolution's weights. Direct convolution doesn't need reshape weights
if(!_is_direct_conv)
{
+ auto w21_tensor = std::unique_ptr<TensorType>(new TensorType());
+ auto w22_tensor = std::unique_ptr<TensorType>(new TensorType());
auto w41_tensor = std::unique_ptr<TensorType>(new TensorType());
auto w42_tensor = std::unique_ptr<TensorType>(new TensorType());
auto w51_tensor = std::unique_ptr<TensorType>(new TensorType());
auto w52_tensor = std::unique_ptr<TensorType>(new TensorType());
+ w21_tensor->allocator()->init(TensorInfo(TensorShape(1248U * data_type_size, 128U / data_type_size), 1, _data_type, _fixed_point_position));
+ w22_tensor->allocator()->init(TensorInfo(TensorShape(1248U * data_type_size, 128U / data_type_size), 1, _data_type, _fixed_point_position));
w41_tensor->allocator()->init(TensorInfo(TensorShape(1920U * data_type_size, 192U / data_type_size), 1, _data_type, _fixed_point_position));
w42_tensor->allocator()->init(TensorInfo(TensorShape(1920U * data_type_size, 192U / data_type_size), 1, _data_type, _fixed_point_position));
w51_tensor->allocator()->init(TensorInfo(TensorShape(1920U * data_type_size, 128U / data_type_size), 1, _data_type, _fixed_point_position));
w52_tensor->allocator()->init(TensorInfo(TensorShape(1920U * data_type_size, 128U / data_type_size), 1, _data_type, _fixed_point_position));
w[2].allocator()->init(TensorInfo(TensorShape(2560U * data_type_size, 384U / data_type_size), 1, _data_type, _fixed_point_position));
+ w21 = std::move(w21_tensor);
+ w22 = std::move(w22_tensor);
w41 = std::move(w41_tensor);
w42 = std::move(w42_tensor);
w51 = std::move(w51_tensor);
@@ -126,12 +125,19 @@
}
else
{
+ w[1].allocator()->init(TensorInfo(TensorShape(5U, 5U, 48U, 256U), 1, _data_type, _fixed_point_position));
+ b[1].allocator()->init(TensorInfo(TensorShape(256U), 1, _data_type, _fixed_point_position));
w[2].allocator()->init(TensorInfo(TensorShape(3U, 3U, 256U, 384U), 1, _data_type, _fixed_point_position));
b[2].allocator()->init(TensorInfo(TensorShape(384U), 1, _data_type, _fixed_point_position));
w[3].allocator()->init(TensorInfo(TensorShape(3U, 3U, 192U, 384U), 1, _data_type, _fixed_point_position));
b[3].allocator()->init(TensorInfo(TensorShape(384U), 1, _data_type, _fixed_point_position));
w[4].allocator()->init(TensorInfo(TensorShape(3U, 3U, 192U, 256U), 1, _data_type, _fixed_point_position));
b[4].allocator()->init(TensorInfo(TensorShape(256U), 1, _data_type, _fixed_point_position));
+ w21 = std::unique_ptr<SubTensorType>(new SubTensorType(&w[1], TensorShape(5U, 5U, 48U, 128U), Coordinates()));
+ w22 = std::unique_ptr<SubTensorType>(new SubTensorType(&w[1], TensorShape(5U, 5U, 48U, 128U), Coordinates(0, 0, 0, 128)));
+ b21 = std::unique_ptr<SubTensorType>(new SubTensorType(&b[1], TensorShape(128U), Coordinates()));
+ b22 = std::unique_ptr<SubTensorType>(new SubTensorType(&b[1], TensorShape(128U), Coordinates(128)));
+
w41 = std::unique_ptr<SubTensorType>(new SubTensorType(&w[3], TensorShape(3U, 3U, 192U, 192U), Coordinates()));
w42 = std::unique_ptr<SubTensorType>(new SubTensorType(&w[3], TensorShape(3U, 3U, 192U, 192U), Coordinates(0, 0, 0, 192)));
b41 = std::unique_ptr<SubTensorType>(new SubTensorType(&b[3], TensorShape(192U), Coordinates()));
@@ -276,10 +282,10 @@
b[6].allocator()->allocate();
b[7].allocator()->allocate();
- dynamic_cast<TensorType *>(w21.get())->allocator()->allocate();
- dynamic_cast<TensorType *>(w22.get())->allocator()->allocate();
if(!_is_direct_conv)
{
+ dynamic_cast<TensorType *>(w21.get())->allocator()->allocate();
+ dynamic_cast<TensorType *>(w22.get())->allocator()->allocate();
dynamic_cast<TensorType *>(w41.get())->allocator()->allocate();
dynamic_cast<TensorType *>(w42.get())->allocator()->allocate();
dynamic_cast<TensorType *>(w51.get())->allocator()->allocate();
@@ -287,9 +293,11 @@
}
else
{
+ b[1].allocator()->allocate();
b[2].allocator()->allocate();
b[3].allocator()->allocate();
b[4].allocator()->allocate();
+ w[1].allocator()->allocate();
w[3].allocator()->allocate();
w[4].allocator()->allocate();
}
@@ -342,11 +350,10 @@
library->fill_tensor_uniform(Accessor(w[7]), 7);
library->fill_tensor_uniform(Accessor(b[7]), 8);
- library->fill_tensor_uniform(Accessor(*dynamic_cast<TensorType *>(w21.get())), 9);
- library->fill_tensor_uniform(Accessor(*dynamic_cast<TensorType *>(w22.get())), 10);
-
if(!_is_direct_conv)
{
+ library->fill_tensor_uniform(Accessor(*dynamic_cast<TensorType *>(w21.get())), 9);
+ library->fill_tensor_uniform(Accessor(*dynamic_cast<TensorType *>(w22.get())), 10);
library->fill_tensor_uniform(Accessor(*dynamic_cast<TensorType *>(w41.get())), 11);
library->fill_tensor_uniform(Accessor(*dynamic_cast<TensorType *>(w42.get())), 12);
library->fill_tensor_uniform(Accessor(*dynamic_cast<TensorType *>(w51.get())), 13);
@@ -354,6 +361,8 @@
}
else
{
+ library->fill_tensor_uniform(Accessor(w[1]), 9);
+ library->fill_tensor_uniform(Accessor(b[1]), 10);
library->fill_tensor_uniform(Accessor(w[3]), 11);
library->fill_tensor_uniform(Accessor(b[3]), 12);
library->fill_tensor_uniform(Accessor(w[4]), 13);
@@ -572,8 +581,8 @@
bool _is_direct_conv{ !std::is_same<ConvolutionLayerFunction, DirectConvolutionLayerFunction>::value };
ActivationLayerFunction act1{}, act2{}, act3{}, act4{}, act5{}, act6{}, act7{};
- ConvolutionLayerFunction conv1{}, conv21{}, conv22{};
- DirectConv conv3{}, conv41{}, conv42{}, conv51{}, conv52{};
+ ConvolutionLayerFunction conv1{};
+ DirectConv conv21{}, conv22{}, conv3{}, conv41{}, conv42{}, conv51{}, conv52{};
FullyConnectedLayerFunction fc6{}, fc7{}, fc8{};
NormalizationLayerFunction norm1{}, norm2{};
PoolingLayerFunction pool1{}, pool2{}, pool5{};