COMPMID-1519: Add support for 3D input/output in CLGEMMLowpOutputStage
Change-Id: I637add70310d2da4d82b236a6352af9d33be17a1
Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/149706
Reviewed-by: Isabella Gottardi <isabella.gottardi@arm.com>
Reviewed-by: Michele DiGiorgio <michele.digiorgio@arm.com>
Tested-by: bsgcomp <bsgcomp@arm.com>
diff --git a/src/runtime/CL/functions/CLGEMMConvolutionLayer.cpp b/src/runtime/CL/functions/CLGEMMConvolutionLayer.cpp
index bd5e969..f41a12a 100644
--- a/src/runtime/CL/functions/CLGEMMConvolutionLayer.cpp
+++ b/src/runtime/CL/functions/CLGEMMConvolutionLayer.cpp
@@ -92,8 +92,8 @@
CLGEMMConvolutionLayer::CLGEMMConvolutionLayer(std::shared_ptr<IMemoryManager> memory_manager)
: _memory_group(memory_manager), _reshape_weights(), _im2col_kernel(), _mm_gemm(memory_manager), _mm_gemmlowp(memory_manager), _gemmlowp_output_stage(), _col2im_kernel(), _activationlayer_function(),
- _add_bias_kernel(), _reshape_layer(), _original_weights(nullptr), _im2col_output(), _weights_reshaped(), _gemm_output(), _tmp_output(), _data_layout(DataLayout::NCHW), _append_bias(false),
- _skip_im2col(false), _is_quantized(false), _is_activationlayer_enabled(false), _is_prepared(false)
+ _add_bias_kernel(), _original_weights(nullptr), _im2col_output(), _weights_reshaped(), _gemm_output(), _tmp_output(), _data_layout(DataLayout::NCHW), _append_bias(false), _skip_im2col(false),
+ _skip_col2im(false), _is_quantized(false), _is_activationlayer_enabled(false), _is_prepared(false)
{
}
@@ -102,6 +102,9 @@
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()
@@ -112,7 +115,7 @@
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, 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);
@@ -121,8 +124,7 @@
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);
}
}
@@ -130,10 +132,11 @@
{
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 /* Reinterpret the input as 3D if im2col is skipped */);
+
if(is_quantized)
{
- const GEMMInfo &gemm_info = GEMMInfo(false, false, true /* Reshape weights only for the first run */);
-
// Since we need negative offsets for computing convolution, we need to change QuantizationInfo()
// Extract and negate input and weights offset
const QuantizationInfo input_quantization_info = input->quantization_info();
@@ -149,8 +152,6 @@
}
else
{
- 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 */);
-
// Perform validation step on Matrix multiply function
return CLGEMM::validate(input, weights, nullptr, output, 1.0f, 0.0f, gemm_info);
}
@@ -175,6 +176,7 @@
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);
@@ -184,14 +186,14 @@
_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) && !_is_quantized;
+ _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);
// Set the GPU target for im2col and col2im
_im2col_kernel.set_target(CLScheduler::get().target());
_col2im_kernel.set_target(CLScheduler::get().target());
- bool is_nhwc = _data_layout == DataLayout::NHWC;
const ICLTensor *gemm_input_to_use = input;
ICLTensor *gemm_output_to_use = output;
ICLTensor *gemm_output_staged_to_use = output;
@@ -241,18 +243,27 @@
}
// Create GEMM output tensor
- if(!is_nhwc || _is_quantized)
+ 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);
-
+ TensorShape shape_gemm;
+ if(_skip_col2im)
+ {
+ shape_gemm = input->info()->tensor_shape();
+ shape_gemm.set(idx_width, conv_w);
+ shape_gemm.set(idx_height, conv_h);
+ shape_gemm.set(idx_channel, mat_weights_cols);
+ }
+ else
+ {
+ 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;
// 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);
@@ -277,30 +288,29 @@
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);
}
- if(!is_nhwc || _is_quantized)
+ if(!_skip_col2im)
{
- if(input->info()->data_layout() == DataLayout::NCHW)
- {
- // Configure and tune Col2Im
- _col2im_kernel.configure(_is_quantized ? gemm_output_staged_to_use : gemm_output_to_use, output, Size2D(conv_w, conv_h), num_groups);
- CLScheduler::get().tune_kernel_static(_col2im_kernel);
- }
- else
- {
- // Configure reshape layer
- _reshape_layer.configure(_is_quantized ? gemm_output_staged_to_use : gemm_output_to_use, output);
- }
+ // Configure and tune Col2Im
+ _col2im_kernel.configure(_is_quantized ? gemm_output_staged_to_use : gemm_output_to_use, output, Size2D(conv_w, conv_h), num_groups);
+ CLScheduler::get().tune_kernel_static(_col2im_kernel);
}
- if(!is_nhwc || _is_quantized)
+ if(!_skip_col2im)
{
_tmp_output.allocator()->allocate();
+ }
+
+ if(!_skip_col2im || _is_quantized)
+ {
_gemm_output.allocator()->allocate();
}
@@ -346,10 +356,10 @@
const ITensorInfo *gemm_output_staged_to_use = output;
const ITensorInfo *weights_to_use = weights;
- const bool is_nhwc = data_layout == DataLayout::NHWC;
const bool is_quantized = is_data_type_quantized_asymmetric(data_type);
- const bool skip_im2col = (data_layout == DataLayout::NHWC && kernel_width == 1 && kernel_height == 1 && conv_info.stride().first == 1 && conv_info.stride().second == 1) && !is_quantized;
const bool append_bias = (biases != nullptr) && (!is_quantized);
+ const bool skip_im2col = (data_layout == DataLayout::NHWC && kernel_width == 1 && kernel_height == 1 && conv_info.stride().first == 1 && conv_info.stride().second == 1);
+ const bool skip_col2im = data_layout == DataLayout::NHWC;
ARM_COMPUTE_RETURN_ERROR_ON((weights->dimension(idx_channel) * num_groups) != input->dimension(idx_channel));
ARM_COMPUTE_RETURN_ERROR_ON(weights->num_dimensions() > 4);
@@ -411,19 +421,30 @@
}
// Create GEMM output tensor
- if(!is_nhwc || is_quantized)
+ if(!skip_col2im || is_quantized)
{
- 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;
+ TensorShape shape_gemm;
+ if(skip_col2im)
+ {
+ shape_gemm = input->tensor_shape();
+ shape_gemm.set(idx_width, conv_w);
+ shape_gemm.set(idx_height, conv_h);
+ shape_gemm.set(idx_channel, mat_weights_cols);
+ }
+ else
+ {
+ shape_gemm = gemm_input_to_use->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.
info_gemm = TensorInfo(shape_gemm, 1, gemm_data_type);
- info_gemm.set_quantization_info(output->quantization_info());
+ 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, (data_layout == DataLayout::NHWC) ? 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 : 1, skip_im2col));
if(is_quantized)
{
@@ -431,23 +452,22 @@
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());
+ gemm_output_staged_to_use = &tmp_info;
+ }
// Validate output stage for quantized case
- CLGEMMLowpQuantizeDownInt32ToUint8ScaleByFixedPoint::validate(gemm_output_to_use, biases, gemm_output_staged_to_use, output->quantization_info().offset);
+ CLGEMMLowpQuantizeDownInt32ToUint8ScaleByFixedPoint::validate(gemm_output_to_use, biases, gemm_output_staged_to_use);
}
// Validate Col2Im
- if(!is_nhwc || is_quantized)
+ if(!skip_col2im)
{
- if(input->data_layout() == DataLayout::NCHW)
- {
- ARM_COMPUTE_RETURN_ON_ERROR(CLCol2ImKernel::validate(is_quantized ? gemm_output_staged_to_use : gemm_output_to_use,
- output,
- Size2D(conv_w, conv_h), num_groups));
- }
+ ARM_COMPUTE_RETURN_ON_ERROR(CLCol2ImKernel::validate(is_quantized ? gemm_output_staged_to_use : gemm_output_to_use, output,
+ Size2D(conv_w, conv_h), num_groups));
}
//Validate Activation Layer
@@ -492,16 +512,9 @@
}
// Reshape output matrix
- if(_data_layout == DataLayout::NCHW || _is_quantized)
+ if(!_skip_col2im)
{
- if(_data_layout == DataLayout::NCHW)
- {
- CLScheduler::get().enqueue(_col2im_kernel, false);
- }
- else
- {
- _reshape_layer.run();
- }
+ CLScheduler::get().enqueue(_col2im_kernel, false);
}
//Run Activation Layer if enabled