arm_compute v18.11
diff --git a/src/runtime/NEON/functions/NEGEMMConvolutionLayer.cpp b/src/runtime/NEON/functions/NEGEMMConvolutionLayer.cpp
index 92e641e..0232a83 100644
--- a/src/runtime/NEON/functions/NEGEMMConvolutionLayer.cpp
+++ b/src/runtime/NEON/functions/NEGEMMConvolutionLayer.cpp
@@ -32,6 +32,7 @@
#include "support/ToolchainSupport.h"
#include <cmath>
+#include <set>
#include <tuple>
using namespace arm_compute;
@@ -100,6 +101,9 @@
ARM_COMPUTE_ERROR_ON_NULLPTR(input, weights);
ARM_COMPUTE_ERROR_THROW_ON(validate_mm(input->info(), weights->info(), output->info(), gemm_3d_depth, _skip_im2col));
+ const GEMMInfo &gemm_info = GEMMInfo(false, false, true /* Reshape weights only for the first run */,
+ gemm_3d_depth, _skip_im2col /* Reinterpret the input as 3D if im2col is skipped */);
+
if(_is_quantized)
{
// Since we need negative offsets for computing convolution, we need to change QuantizationInfo()
@@ -110,7 +114,7 @@
input->info()->set_quantization_info(QuantizationInfo(input_quantization_info.scale, -input_quantization_info.offset));
weights->info()->set_quantization_info(QuantizationInfo(weights_quantization_info.scale, -weights_quantization_info.offset));
- _mm_gemmlowp.configure(input, weights, output, GEMMInfo(false, false, true /* Reshape weights only for the first run*/));
+ _mm_gemmlowp.configure(input, weights, nullptr, output, gemm_info);
// Revert back QuantizatioInfo as input and weights could be used in other convolution layers
input->info()->set_quantization_info(input_quantization_info);
@@ -119,8 +123,7 @@
else
{
// Configure matrix multiply function
- _mm_gemm.configure(input, weights, nullptr, output, 1.0f, 0.0f, GEMMInfo(false, false, true /* Reshape weights only for the first run*/, gemm_3d_depth,
- _skip_im2col /* Reinterpret the input as 3D if im2col is skipped */));
+ _mm_gemm.configure(input, weights, nullptr, output, 1.0f, 0.0f, gemm_info);
}
}
@@ -128,7 +131,8 @@
{
const bool is_quantized = is_data_type_quantized_asymmetric(input->data_type());
- const GEMMInfo gemm_info = GEMMInfo(false, false, true /* Reshape weights only for the first run */, gemm_3d_depth, skip_im2col);
+ const GEMMInfo &gemm_info = GEMMInfo(false, false, true /* Reshape weights only for the first run */,
+ gemm_3d_depth, skip_im2col /* Reinterpret the input as 3D if im2col is skipped */);
if(is_quantized)
{
// Since we need negative offsets for computing convolution, we need to change QuantizationInfo()
@@ -142,7 +146,7 @@
weights_qa->set_quantization_info(QuantizationInfo(weights_quantization_info.scale, -weights_quantization_info.offset));
// Perform validation step on GEMMLowp
- return NEGEMMLowpMatrixMultiplyCore::validate(input_qa.get(), weights_qa.get(), output, gemm_info);
+ return NEGEMMLowpMatrixMultiplyCore::validate(input_qa.get(), weights_qa.get(), nullptr, output, gemm_info);
}
else
{
@@ -185,19 +189,18 @@
const DataLayout data_layout = input->info()->data_layout();
const int idx_width = get_data_layout_dimension_index(data_layout, DataLayoutDimension::WIDTH);
const int idx_height = get_data_layout_dimension_index(data_layout, DataLayoutDimension::HEIGHT);
- const int idx_channel = get_data_layout_dimension_index(data_layout, DataLayoutDimension::CHANNEL);
const int idx_kernels = get_data_layout_dimension_index(data_layout, DataLayoutDimension::BATCHES);
const unsigned int kernel_width = weights->info()->dimension(idx_width);
const unsigned int kernel_height = weights->info()->dimension(idx_height);
- _is_prepared = weights_info.retain_internal_weights();
- _original_weights = weights;
- _is_quantized = is_data_type_quantized_asymmetric(input->info()->data_type());
- _data_layout = data_layout;
- _skip_im2col = (data_layout == DataLayout::NHWC && kernel_width == 1 && kernel_height == 1 && conv_info.stride().first == 1 && conv_info.stride().second == 1);
- _skip_col2im = data_layout == DataLayout::NHWC;
- _append_bias = (biases != nullptr) && (!_is_quantized);
+ _is_prepared = weights_info.retain_internal_weights();
+ _original_weights = weights;
+ _is_quantized = is_data_type_quantized_asymmetric(input->info()->data_type());
+ _data_layout = data_layout;
+ _skip_im2col = (data_layout == DataLayout::NHWC && kernel_width == 1 && kernel_height == 1 && conv_info.stride().first == 1 && conv_info.stride().second == 1);
+ _append_bias = (biases != nullptr) && (!_is_quantized);
+ _is_activationlayer_enabled = act_info.enabled();
const ITensor *gemm_input_to_use = input;
ITensor *gemm_output_to_use = output;
@@ -214,17 +217,20 @@
dilation);
// Check if GEMM3D is supported
- if(_skip_col2im)
+ if(data_layout == DataLayout::NHWC)
{
+ _skip_col2im = bool(validate_gemm3d(input->info()->data_type(), conv_h, true));
// If not supported, we need to perform im2col and col2im (or reshape layer)
- if(!bool(validate_gemm3d(input->info()->data_type(), conv_h, _skip_im2col)))
+ if(!_skip_col2im)
{
_skip_im2col = false;
- _skip_col2im = false;
}
}
+ else
+ {
+ _skip_col2im = false;
+ }
- const unsigned bias_element = (_append_bias && !_skip_im2col) ? 1 : 0;
const ITensor *biases_to_use = (_append_bias && !_skip_im2col) ? biases : nullptr;
// Get parameters from conv_info
@@ -233,7 +239,6 @@
std::tie(stride_x, stride_y) = conv_info.stride();
unsigned int mat_weights_cols = weights->info()->dimension(idx_kernels);
- unsigned int mat_weights_rows = weights->info()->dimension(idx_width) * weights->info()->dimension(idx_height) * weights->info()->dimension(idx_channel) + bias_element;
// _weights_reshaped will be auto configured in the kernel.
// Just append biases and do not transpose 1xW as it will be reshaped in NEGEMM
@@ -242,14 +247,6 @@
// Create tensor to store im2col reshaped inputs
if(!_skip_im2col)
{
- // Calculate im2col shape
- // For NEON the batch size is on the fourth dimension
- TensorShape shape_im2col = input->info()->tensor_shape();
- shape_im2col.set(0, mat_weights_rows);
- shape_im2col.set(1, conv_w * conv_h);
- shape_im2col.set(2, 1);
-
- _im2col_output.allocator()->init(input->info()->clone()->set_is_resizable(true).reset_padding().set_tensor_shape(shape_im2col));
_memory_group.manage(&_im2col_output);
// Configure
@@ -265,17 +262,27 @@
}
// Create temporary GEMM output tensor in case we cannot skip col2im
- if(!_skip_col2im)
+ if(!_skip_col2im || _is_quantized)
{
- // Calculate GEMM output shape
- TensorShape shape_gemm = _im2col_output.info()->tensor_shape();
- shape_gemm.set(0, mat_weights_cols);
- shape_gemm.set(1, conv_w * conv_h);
-
// GEMM output should be S32 for acquiring raw integer accumulator without quantized postprocessing for quantized asymmetric input.
const DataType gemm_data_type = _is_quantized ? DataType::S32 : data_type;
+ TensorShape shape_gemm;
+
+ if(_is_quantized && _skip_col2im)
+ {
+ shape_gemm = output->info()->tensor_shape();
+ }
+ else
+ {
+ // Calculate GEMM output shape
+ shape_gemm = _im2col_output.info()->tensor_shape();
+ shape_gemm.set(0, mat_weights_cols);
+ shape_gemm.set(1, conv_w * conv_h);
+ }
+
+ // FIXME: input->clone() doesn't work with subtensors for grouped convolutions.
TensorInfo info_gemm(shape_gemm, 1, gemm_data_type);
- info_gemm.set_quantization_info(output->info()->quantization_info());
+ info_gemm.set_quantization_info(output->info()->quantization_info()).set_data_layout(input->info()->data_layout());
_gemm_output.allocator()->init(info_gemm);
_memory_group.manage(&_gemm_output);
@@ -284,7 +291,9 @@
}
// Configure GEMM
- configure_mm(gemm_input_to_use, &_weights_reshaped, gemm_output_to_use, _skip_col2im ? conv_h : 1);
+ // In case we need to skip col2im, GEMM3D (gemm_3d_depth != 0) must be called in order to avoid reshaping the output matrix
+ const unsigned int gemm_3d_depth = _skip_col2im ? conv_h : 0;
+ configure_mm(gemm_input_to_use, &_weights_reshaped, gemm_output_to_use, gemm_3d_depth);
if(!_skip_im2col)
{
@@ -294,16 +303,39 @@
// Configure output stage for quantized case
if(_is_quantized)
{
- const QuantizationInfo output_quant_info = (output->info()->total_size() == 0) ? input->info()->quantization_info() : output->info()->quantization_info();
+ const QuantizationInfo input_quant_info = input->info()->quantization_info();
+ const QuantizationInfo output_quant_info = (output->info()->total_size() == 0) ? input_quant_info : output->info()->quantization_info();
- float multiplier = input->info()->quantization_info().scale * weights->info()->quantization_info().scale / output_quant_info.scale;
+ float multiplier = input_quant_info.scale * weights->info()->quantization_info().scale / output_quant_info.scale;
int output_multiplier, output_shift;
quantization::calculate_quantized_multiplier_less_than_one(multiplier, &output_multiplier, &output_shift);
- _memory_group.manage(&_tmp_output);
- gemm_output_staged_to_use = &_tmp_output;
+ if(!_skip_col2im)
+ {
+ _memory_group.manage(&_tmp_output);
+ gemm_output_staged_to_use = &_tmp_output;
+ }
- _gemmlowp_output_stage.configure(gemm_output_to_use, biases, gemm_output_staged_to_use, output_multiplier, output_shift, output_quant_info.offset);
+ // Merge activation with output stage
+ int min_activation = 0;
+ int max_activation = 0;
+
+ const std::set<ActivationLayerInfo::ActivationFunction> supported_acts = { ActivationLayerInfo::ActivationFunction::RELU,
+ ActivationLayerInfo::ActivationFunction::BOUNDED_RELU,
+ ActivationLayerInfo::ActivationFunction::LU_BOUNDED_RELU
+ };
+ if(_is_activationlayer_enabled && supported_acts.count(act_info.activation()) != 0)
+ {
+ const int a_const_int = output_quant_info.quantize(act_info.a(), RoundingPolicy::TO_NEAREST_UP);
+ const int b_const_int = output_quant_info.quantize(act_info.b(), RoundingPolicy::TO_NEAREST_UP);
+
+ min_activation = act_info.activation() != ActivationLayerInfo::ActivationFunction::LU_BOUNDED_RELU ? output_quant_info.offset : b_const_int;
+ max_activation = act_info.activation() == ActivationLayerInfo::ActivationFunction::RELU ? 255 : a_const_int;
+
+ _is_activationlayer_enabled = false;
+ }
+
+ _gemmlowp_output_stage.configure(gemm_output_to_use, biases, gemm_output_staged_to_use, output_multiplier, output_shift, output_quant_info.offset, min_activation, max_activation);
}
if(!_skip_col2im)
@@ -320,12 +352,12 @@
}
}
- if(_is_quantized)
+ if(_is_quantized && !_skip_col2im)
{
_tmp_output.allocator()->allocate();
}
- if(!_skip_col2im)
+ if(!_skip_col2im || _is_quantized)
{
_gemm_output.allocator()->allocate();
}
@@ -333,9 +365,7 @@
ARM_COMPUTE_ERROR_ON_MSG((output->info()->dimension(idx_width) != conv_w) || (output->info()->dimension(idx_height) != conv_h),
"Output shape does not match the expected one");
- //Configure Activation Layer
- _is_activationlayer_enabled = act_info.enabled();
-
+ // Configure Activation Layer
if(_is_activationlayer_enabled)
{
_activationlayer_function.configure(output, nullptr, act_info);
@@ -370,10 +400,10 @@
const ITensorInfo *gemm_output_staged_to_use = output;
const ITensorInfo *weights_to_use = weights;
- const bool is_quantized = is_data_type_quantized_asymmetric(data_type);
- const bool append_bias = (biases != nullptr) && (!is_quantized);
- bool skip_im2col = (data_layout == DataLayout::NHWC && kernel_width == 1 && kernel_height == 1 && conv_info.stride().first == 1 && conv_info.stride().second == 1);
- bool skip_col2im = data_layout == DataLayout::NHWC;
+ const bool is_quantized = is_data_type_quantized_asymmetric(data_type);
+ const bool append_bias = (biases != nullptr) && (!is_quantized);
+ bool skip_im2col = (data_layout == DataLayout::NHWC && kernel_width == 1 && kernel_height == 1 && conv_info.stride().first == 1 && conv_info.stride().second == 1);
+ bool is_activation_enabled = act_info.enabled();
// Get convolved dimensions
unsigned int conv_w = 0;
@@ -387,6 +417,17 @@
dilation);
// Check if GEMM3D is supported
+ bool skip_col2im = false;
+ if(data_layout == DataLayout::NHWC)
+ {
+ skip_col2im = bool(validate_gemm3d(input->data_type(), conv_h, true));
+ // If not supported, we need to perform im2col and col2im (or reshape layer)
+ if(!skip_col2im)
+ {
+ skip_im2col = false;
+ }
+ }
+
if(skip_col2im)
{
// If not supported, we need to perform im2col and col2im (or reshape layer)
@@ -435,6 +476,7 @@
{
// Create tensor info for im2col reshaped inputs
// For NEON the batch size is on the fourth dimension
+ // TODO (giaiod01): Auto-initialize the output shape of im2col COMPMID-1482
TensorShape shape_im2col = input->tensor_shape();
shape_im2col.set(0, mat_weights_rows);
shape_im2col.set(1, conv_w * conv_h);
@@ -453,33 +495,60 @@
}
// Create temporary GEMM output tensor in case we cannot skip col2im
+ const DataType gemm_data_type = is_quantized ? DataType::S32 : data_type;
if(!skip_col2im)
{
TensorShape shape_gemm = gemm_input_to_use->tensor_shape();
shape_gemm.set(0, mat_weights_cols);
shape_gemm.set(1, conv_w * conv_h);
- const DataType gemm_data_type = is_quantized ? DataType::S32 : data_type;
- // GEMM output should be S32 for acquiring raw integer accumulator without quantized postprocessing for quantized asymmetric input.
info_gemm = TensorInfo(shape_gemm, 1, gemm_data_type);
- info_gemm.set_quantization_info(output->quantization_info());
-
- gemm_output_to_use = &info_gemm;
}
+ else
+ {
+ info_gemm = TensorInfo(output->tensor_shape(), 1, gemm_data_type);
+ }
+ info_gemm.set_quantization_info(output->quantization_info()).set_data_layout(input->data_layout());
+ gemm_output_to_use = &info_gemm;
- ARM_COMPUTE_RETURN_ON_ERROR(validate_mm(gemm_input_to_use, weights_to_use, gemm_output_to_use, skip_col2im ? conv_h : 1, skip_im2col));
+ ARM_COMPUTE_RETURN_ON_ERROR(validate_mm(gemm_input_to_use, weights_to_use, gemm_output_to_use, skip_col2im ? conv_h : 0, skip_im2col));
if(is_quantized)
{
- float multiplier = input->quantization_info().scale * weights_to_use->quantization_info().scale / output->quantization_info().scale;
- int output_multiplier, output_shift;
+ const QuantizationInfo input_quant_info = input->quantization_info();
+ const QuantizationInfo output_quant_info = (output->total_size() == 0) ? input_quant_info : output->quantization_info();
+ const float multiplier = input_quant_info.scale * weights_to_use->quantization_info().scale / output_quant_info.scale;
+ int output_multiplier, output_shift;
quantization::calculate_quantized_multiplier_less_than_one(multiplier, &output_multiplier, &output_shift);
- tmp_info = TensorInfo(gemm_output_to_use->tensor_shape(), 1, DataType::QASYMM8);
- tmp_info.set_quantization_info(output->quantization_info());
- gemm_output_staged_to_use = &tmp_info;
+ if(!skip_col2im)
+ {
+ tmp_info = TensorInfo(gemm_output_to_use->tensor_shape(), 1, DataType::QASYMM8);
+ tmp_info.set_quantization_info(output->quantization_info()).set_data_layout(data_layout);
+ gemm_output_staged_to_use = &tmp_info;
+ }
+
+ // Merge activation with output stage
+ int min_activation = 0;
+ int max_activation = 0;
+
+ const std::set<ActivationLayerInfo::ActivationFunction> supported_acts = { ActivationLayerInfo::ActivationFunction::RELU,
+ ActivationLayerInfo::ActivationFunction::BOUNDED_RELU,
+ ActivationLayerInfo::ActivationFunction::LU_BOUNDED_RELU
+ };
+
+ if(is_activation_enabled && supported_acts.count(act_info.activation()) != 0)
+ {
+ const int a_const_int = output_quant_info.quantize(act_info.a(), RoundingPolicy::TO_NEAREST_UP);
+ const int b_const_int = output_quant_info.quantize(act_info.b(), RoundingPolicy::TO_NEAREST_UP);
+
+ min_activation = act_info.activation() != ActivationLayerInfo::ActivationFunction::LU_BOUNDED_RELU ? output_quant_info.offset : b_const_int;
+ max_activation = act_info.activation() == ActivationLayerInfo::ActivationFunction::RELU ? 255 : a_const_int;
+
+ is_activation_enabled = false;
+ }
// Validate output stage for quantized case
- NEGEMMLowpQuantizeDownInt32ToUint8ScaleByFixedPoint::validate(gemm_output_to_use, biases, gemm_output_staged_to_use, output->quantization_info().offset);
+ NEGEMMLowpQuantizeDownInt32ToUint8ScaleByFixedPoint::validate(gemm_output_to_use, biases, gemm_output_staged_to_use, min_activation, max_activation);
}
// Validate Col2Im/ReshapeLayer
@@ -491,7 +560,7 @@
}
//Validate Activation Layer
- if(act_info.enabled())
+ if(is_activation_enabled)
{
ARM_COMPUTE_RETURN_ON_ERROR(NEActivationLayer::validate(output, nullptr, act_info));
}