Fixed misleading test case bug in deconv1d (#633)
Correct erros of deconv1d unittest Added hint in the header indicating that padtype is not supported yet Added 2 cases for deconv1d Type: Code Improvement Issue: github issue #585 Signed-off-by: Feiyue Chen <Feiyue.Chen@verisilicon.com>
This commit is contained in:
Chen Feiyue
GitHub
parent
cf2efc63fd
commit
9bb3e7c68b
|
|
@ -42,53 +42,49 @@ namespace ops {
|
|||
*
|
||||
* - weights : the channel number for weight tensor.
|
||||
* - ksize : the length for weight tensor.
|
||||
* - padding : AUTO, VALID or SAME.
|
||||
* - padtype : AUTO, VALID or SAME.**
|
||||
* - pad : pad value for each spatial axis.
|
||||
* - stride : stride along each spatial axis.
|
||||
* - output_padding : specifying the amount of padding along the height and width of
|
||||
* the output tensor.
|
||||
* - output_padding : additional padding lines added to the output tensor, default is zero
|
||||
*
|
||||
* Caution**: PadType is not really supported yet, will be supported in future.
|
||||
*/
|
||||
|
||||
class DeConv1d : public BuiltinOp {
|
||||
public:
|
||||
DeConv1d(Graph* graph, PadType pad_type,
|
||||
uint32_t stride, uint32_t output_padding, uint32_t group = 1,
|
||||
DataLayout input_layout = DataLayout::WHCN,
|
||||
DataLayout kernel_layout = DataLayout::WHIcOc);
|
||||
DeConv1d(Graph* graph, const std::array<uint32_t, 2>& pad,
|
||||
uint32_t stride, uint32_t output_padding, uint32_t group = 1,
|
||||
DataLayout input_layout = DataLayout::WHCN,
|
||||
DataLayout kernel_layout = DataLayout::WHIcOc);
|
||||
DeConv1d(Graph* graph, int32_t oc_count_, PadType pad_type,
|
||||
uint32_t ksize,
|
||||
uint32_t stride,
|
||||
uint32_t output_padding);
|
||||
DeConv1d(Graph* graph, int32_t oc_count_, PadType pad_type,
|
||||
uint32_t ksize,
|
||||
uint32_t stride,
|
||||
uint32_t output_padding,
|
||||
const std::array<uint32_t, 2>& pad,
|
||||
uint32_t group = 1);
|
||||
DeConv1d(Graph* graph, PadType pad_type,
|
||||
uint32_t stride, uint32_t output_padding,
|
||||
const std::array<uint32_t, 2>& pad, uint32_t group,
|
||||
DataLayout input_layout, DataLayout kernel_layout);
|
||||
public:
|
||||
DeConv1d(Graph* graph, PadType pad_type, uint32_t stride,
|
||||
uint32_t output_padding, uint32_t group = 1,
|
||||
DataLayout input_layout = DataLayout::WHCN,
|
||||
DataLayout kernel_layout = DataLayout::WHIcOc);
|
||||
DeConv1d(Graph* graph, const std::array<uint32_t, 2>& pad, uint32_t stride,
|
||||
uint32_t output_padding, uint32_t group = 1,
|
||||
DataLayout input_layout = DataLayout::WHCN,
|
||||
DataLayout kernel_layout = DataLayout::WHIcOc);
|
||||
DeConv1d(Graph* graph, int32_t oc_count_, PadType pad_type, uint32_t ksize,
|
||||
uint32_t stride, uint32_t output_padding);
|
||||
DeConv1d(Graph* graph, int32_t oc_count_, PadType pad_type, uint32_t ksize,
|
||||
uint32_t stride, uint32_t output_padding,
|
||||
const std::array<uint32_t, 2>& pad, uint32_t group = 1);
|
||||
DeConv1d(Graph* graph, PadType pad_type, uint32_t stride,
|
||||
uint32_t output_padding, const std::array<uint32_t, 2>& pad,
|
||||
uint32_t group, DataLayout input_layout, DataLayout kernel_layout);
|
||||
|
||||
std::shared_ptr<Operation> Clone(std::shared_ptr<Graph>& graph) const override;
|
||||
std::shared_ptr<Operation> Clone(
|
||||
std::shared_ptr<Graph>& graph) const override;
|
||||
|
||||
protected:
|
||||
const uint32_t oc_count_; // output channel count
|
||||
const PadType pad_type_;
|
||||
const uint32_t ksize_;
|
||||
const uint32_t stride_;
|
||||
const uint32_t output_padding_;
|
||||
const std::array<uint32_t, 2> pad_;
|
||||
const uint32_t group_;
|
||||
const DataLayout kernel_layout_;
|
||||
protected:
|
||||
const uint32_t oc_count_; // output channel count
|
||||
const PadType pad_type_;
|
||||
const uint32_t ksize_;
|
||||
const uint32_t stride_;
|
||||
const uint32_t output_padding_;
|
||||
const std::array<uint32_t, 2> pad_;
|
||||
const uint32_t group_;
|
||||
const DataLayout kernel_layout_;
|
||||
};
|
||||
|
||||
} // namespace ops
|
||||
} // namespace vx
|
||||
} // namespace tim
|
||||
} // namespace ops
|
||||
} // namespace vx
|
||||
} // namespace tim
|
||||
|
||||
#endif /* TIM_VX_OPS_DECONV1D_H_ */
|
||||
|
|
|
|||
|
|
@ -24,110 +24,238 @@
|
|||
#include "tim/vx/context.h"
|
||||
#include "tim/vx/graph.h"
|
||||
#include "tim/vx/ops/deconv1d.h"
|
||||
#include "tim/vx/ops/activations.h"
|
||||
|
||||
#include "gtest/gtest.h"
|
||||
|
||||
TEST(DeConv1d, no_bias_layout_whcn_depthwise_shape_3_2_1) {
|
||||
auto ctx = tim::vx::Context::Create();
|
||||
auto graph = ctx->CreateGraph();
|
||||
TEST(DeConv1d, no_bias_no_outputpadding_shape_3_2_1) {
|
||||
auto ctx = tim::vx::Context::Create();
|
||||
auto graph = ctx->CreateGraph();
|
||||
|
||||
tim::vx::ShapeType input_shape ({3, 2, 1}); //whcn
|
||||
tim::vx::ShapeType kernel_shape({3, 2, 1}); //whc1 same as depthwise convolution
|
||||
tim::vx::ShapeType output_shape({5, 2, 1}); //whcn
|
||||
tim::vx::ShapeType input_shape({3, 2, 1});
|
||||
tim::vx::ShapeType kernel_shape({3, 2, 1});
|
||||
tim::vx::ShapeType output_shape({5, 1, 1});
|
||||
|
||||
tim::vx::TensorSpec input_spec (tim::vx::DataType::FLOAT32, input_shape, tim::vx::TensorAttribute::INPUT);
|
||||
tim::vx::TensorSpec kernel_spec (tim::vx::DataType::FLOAT32, kernel_shape, tim::vx::TensorAttribute::CONSTANT);
|
||||
tim::vx::TensorSpec output_spec (tim::vx::DataType::FLOAT32, output_shape, tim::vx::TensorAttribute::OUTPUT);
|
||||
tim::vx::TensorSpec input_spec(tim::vx::DataType::FLOAT32, input_shape,
|
||||
tim::vx::TensorAttribute::INPUT);
|
||||
tim::vx::TensorSpec kernel_spec(tim::vx::DataType::FLOAT32, kernel_shape,
|
||||
tim::vx::TensorAttribute::CONSTANT);
|
||||
tim::vx::TensorSpec output_spec(tim::vx::DataType::FLOAT32, output_shape,
|
||||
tim::vx::TensorAttribute::OUTPUT);
|
||||
|
||||
auto input_tensor = graph->CreateTensor(input_spec);
|
||||
auto output_tensor = graph->CreateTensor(output_spec);
|
||||
auto kernel_tensor = graph->CreateTensor(kernel_spec);
|
||||
auto input_tensor = graph->CreateTensor(input_spec);
|
||||
auto output_tensor = graph->CreateTensor(output_spec);
|
||||
auto kernel_tensor = graph->CreateTensor(kernel_spec);
|
||||
|
||||
std::vector<float> input_data = { 3.0f, 9.0f, 3.0f,
|
||||
7.0f, 5.0f, 9.0f, };
|
||||
std::vector<float> kernel_data = { 9.0f, 0.0f, 1.0f,
|
||||
3.0f, 0.0f, 0.0f, };
|
||||
std::vector<float> input_data = {
|
||||
3.0f, 9.0f, 3.0f, 7.0f, 5.0f, 9.0f,
|
||||
};
|
||||
std::vector<float> kernel_data = {
|
||||
9.0f, 0.0f, 1.0f, 3.0f, 0.0f, 0.0f,
|
||||
};
|
||||
|
||||
std::vector<float> golden = {
|
||||
27.0f, 81.0f, 30.0f, 9.0f, 3.0f,
|
||||
21.0f, 15.0f, 27.0f, 0.0f, 0.0f, };
|
||||
std::vector<float> golden = {48, 96, 57, 9, 3};
|
||||
|
||||
std::vector<float> output_data(golden.size());
|
||||
std::vector<float> output_data(golden.size());
|
||||
|
||||
EXPECT_TRUE(input_tensor->CopyDataToTensor(input_data.data(), input_data.size()*4));
|
||||
EXPECT_TRUE(kernel_tensor->CopyDataToTensor(kernel_data.data(), kernel_data.size()*4));
|
||||
EXPECT_TRUE(
|
||||
input_tensor->CopyDataToTensor(input_data.data(), input_data.size() * 4));
|
||||
EXPECT_TRUE(kernel_tensor->CopyDataToTensor(kernel_data.data(),
|
||||
kernel_data.size() * 4));
|
||||
|
||||
auto op = graph->CreateOperation<tim::vx::ops::DeConv1d>(
|
||||
2, tim::vx::PadType::SAME, 3, 1, 1, std::array<uint32_t, 2>({0, 0}), 2);
|
||||
(*op).BindInputs({input_tensor, kernel_tensor}).BindOutputs({output_tensor});
|
||||
auto op = graph->CreateOperation<tim::vx::ops::DeConv1d>(
|
||||
2, tim::vx::PadType::VALID, /*ksize=*/3, /*stride=*/1,
|
||||
/*output_padding=*/0, std::array<uint32_t, 2>({0, 0}), 1);
|
||||
(*op).BindInputs({input_tensor, kernel_tensor}).BindOutputs({output_tensor});
|
||||
|
||||
EXPECT_TRUE(graph->Compile());
|
||||
EXPECT_TRUE(graph->Run());
|
||||
EXPECT_TRUE(graph->Compile());
|
||||
EXPECT_TRUE(graph->Run());
|
||||
|
||||
EXPECT_TRUE(output_tensor->CopyDataFromTensor(output_data.data()));
|
||||
EXPECT_TRUE(output_tensor->CopyDataFromTensor(output_data.data()));
|
||||
|
||||
EXPECT_EQ(golden, output_data) << "Result mismatch";
|
||||
EXPECT_EQ(golden, output_data) << "Result mismatch";
|
||||
}
|
||||
|
||||
TEST(DeConv1d, layout_whcn_shape_3_1_1) {
|
||||
auto ctx = tim::vx::Context::Create();
|
||||
auto graph = ctx->CreateGraph();
|
||||
TEST(DeConv1d, no_bias_has_outputpadding_shape_3_2_1) {
|
||||
auto ctx = tim::vx::Context::Create();
|
||||
auto graph = ctx->CreateGraph();
|
||||
|
||||
tim::vx::ShapeType input_shape ({3, 1, 1});
|
||||
tim::vx::ShapeType kernel_shape({3, 1, 1});
|
||||
tim::vx::ShapeType output_shape({5, 1, 1});
|
||||
tim::vx::ShapeType bias_shape({1});
|
||||
tim::vx::ShapeType input_shape({3, 2, 1});
|
||||
tim::vx::ShapeType kernel_shape({3, 2, 1});
|
||||
tim::vx::ShapeType output_shape({6, 1, 1});
|
||||
|
||||
tim::vx::Quantization input_quant(tim::vx::QuantType::ASYMMETRIC, 1.0f, 0);
|
||||
tim::vx::Quantization output_quant(tim::vx::QuantType::ASYMMETRIC, 1.0f, 2);
|
||||
tim::vx::Quantization weight_quant(tim::vx::QuantType::ASYMMETRIC, 1.0f, 0);
|
||||
tim::vx::Quantization bias_quant(tim::vx::QuantType::ASYMMETRIC, 1.0f, 0);
|
||||
tim::vx::TensorSpec input_spec(tim::vx::DataType::FLOAT32, input_shape,
|
||||
tim::vx::TensorAttribute::INPUT);
|
||||
tim::vx::TensorSpec kernel_spec(tim::vx::DataType::FLOAT32, kernel_shape,
|
||||
tim::vx::TensorAttribute::CONSTANT);
|
||||
tim::vx::TensorSpec output_spec(tim::vx::DataType::FLOAT32, output_shape,
|
||||
tim::vx::TensorAttribute::OUTPUT);
|
||||
|
||||
tim::vx::TensorSpec input_spec (
|
||||
tim::vx::DataType::UINT8, input_shape, tim::vx::TensorAttribute::INPUT, input_quant);
|
||||
tim::vx::TensorSpec kernel_spec (
|
||||
tim::vx::DataType::UINT8, kernel_shape, tim::vx::TensorAttribute::CONSTANT, weight_quant);
|
||||
tim::vx::TensorSpec bias_spec (
|
||||
tim::vx::DataType::INT32, bias_shape, tim::vx::TensorAttribute::CONSTANT, bias_quant);
|
||||
tim::vx::TensorSpec output_spec (
|
||||
tim::vx::DataType::UINT8, output_shape, tim::vx::TensorAttribute::OUTPUT, output_quant);
|
||||
auto input_tensor = graph->CreateTensor(input_spec);
|
||||
auto output_tensor = graph->CreateTensor(output_spec);
|
||||
auto kernel_tensor = graph->CreateTensor(kernel_spec);
|
||||
|
||||
auto input_tensor = graph->CreateTensor(input_spec);
|
||||
auto output_tensor = graph->CreateTensor(output_spec);
|
||||
auto kernel_tensor = graph->CreateTensor(kernel_spec);
|
||||
auto bias_tensor = graph->CreateTensor(bias_spec);
|
||||
std::vector<float> input_data = {
|
||||
3.0f, 9.0f, 3.0f, 7.0f, 5.0f, 9.0f,
|
||||
};
|
||||
std::vector<float> kernel_data = {
|
||||
9.0f, 0.0f, 1.0f, 3.0f, 0.0f, 0.0f,
|
||||
};
|
||||
|
||||
std::vector<uint8_t> input_data = {
|
||||
3, 9, 3,
|
||||
};
|
||||
std::vector<float> golden = {48, 96, 57, 9, 3, 0};
|
||||
|
||||
std::vector<uint8_t> kernel_data = {
|
||||
9, 0, 1,
|
||||
};
|
||||
std::vector<float> output_data(golden.size());
|
||||
|
||||
std::vector<int32_t> bias_data = {
|
||||
-5,
|
||||
};
|
||||
EXPECT_TRUE(
|
||||
input_tensor->CopyDataToTensor(input_data.data(), input_data.size() * 4));
|
||||
EXPECT_TRUE(kernel_tensor->CopyDataToTensor(kernel_data.data(),
|
||||
kernel_data.size() * 4));
|
||||
|
||||
std::vector<uint8_t> golden = {
|
||||
24, 78, 27, 6, 0,
|
||||
};
|
||||
auto op = graph->CreateOperation<tim::vx::ops::DeConv1d>(
|
||||
2, tim::vx::PadType::VALID, /*ksize=*/3, /*stride=*/1,
|
||||
/*output_padding=*/ 1, std::array<uint32_t, 2>({0, 0}), 1);
|
||||
(*op).BindInputs({input_tensor, kernel_tensor}).BindOutputs({output_tensor});
|
||||
|
||||
std::vector<uint8_t> output_data(golden.size());
|
||||
EXPECT_TRUE(graph->Compile());
|
||||
EXPECT_TRUE(graph->Run());
|
||||
|
||||
EXPECT_TRUE(input_tensor->CopyDataToTensor(input_data.data(), input_data.size()));
|
||||
EXPECT_TRUE(kernel_tensor->CopyDataToTensor(kernel_data.data(), kernel_data.size()));
|
||||
EXPECT_TRUE(bias_tensor->CopyDataToTensor(bias_data.data(), bias_data.size() * sizeof(int32_t)));
|
||||
EXPECT_TRUE(output_tensor->CopyDataFromTensor(output_data.data()));
|
||||
|
||||
auto op = graph->CreateOperation<tim::vx::ops::DeConv1d>(
|
||||
1, tim::vx::PadType::SAME, 3, 1, 1,
|
||||
std::array<uint32_t, 2>({0, 0,}),
|
||||
1);
|
||||
(*op).BindInputs({input_tensor, kernel_tensor, bias_tensor}).BindOutputs({output_tensor});
|
||||
EXPECT_EQ(golden, output_data) << "Result mismatch";
|
||||
}
|
||||
|
||||
EXPECT_TRUE(graph->Compile());
|
||||
EXPECT_TRUE(graph->Run());
|
||||
TEST(DeConv1d, layout_wcn_shape_3_1_1) {
|
||||
auto ctx = tim::vx::Context::Create();
|
||||
auto graph = ctx->CreateGraph();
|
||||
|
||||
EXPECT_TRUE(output_tensor->CopyDataFromTensor(output_data.data()));
|
||||
EXPECT_EQ(golden, output_data) << "Result mismatch";
|
||||
tim::vx::ShapeType input_shape({3, 1, 1});
|
||||
tim::vx::ShapeType kernel_shape({3, 1, 1});
|
||||
tim::vx::ShapeType output_shape({5, 1, 1});
|
||||
tim::vx::ShapeType bias_shape({1});
|
||||
|
||||
tim::vx::Quantization input_quant(tim::vx::QuantType::ASYMMETRIC, 1.0f, 0);
|
||||
tim::vx::Quantization output_quant(tim::vx::QuantType::ASYMMETRIC, 1.0f, 2);
|
||||
tim::vx::Quantization weight_quant(tim::vx::QuantType::ASYMMETRIC, 1.0f, 0);
|
||||
tim::vx::Quantization bias_quant(tim::vx::QuantType::ASYMMETRIC, 1.0f, 0);
|
||||
|
||||
tim::vx::TensorSpec input_spec(tim::vx::DataType::UINT8, input_shape,
|
||||
tim::vx::TensorAttribute::INPUT, input_quant);
|
||||
tim::vx::TensorSpec kernel_spec(tim::vx::DataType::UINT8, kernel_shape,
|
||||
tim::vx::TensorAttribute::CONSTANT,
|
||||
weight_quant);
|
||||
tim::vx::TensorSpec bias_spec(tim::vx::DataType::INT32, bias_shape,
|
||||
tim::vx::TensorAttribute::CONSTANT, bias_quant);
|
||||
tim::vx::TensorSpec output_spec(tim::vx::DataType::UINT8, output_shape,
|
||||
tim::vx::TensorAttribute::OUTPUT,
|
||||
output_quant);
|
||||
|
||||
auto input_tensor = graph->CreateTensor(input_spec);
|
||||
auto output_tensor = graph->CreateTensor(output_spec);
|
||||
auto kernel_tensor = graph->CreateTensor(kernel_spec);
|
||||
auto bias_tensor = graph->CreateTensor(bias_spec);
|
||||
|
||||
std::vector<uint8_t> input_data = {
|
||||
3,
|
||||
9,
|
||||
3,
|
||||
};
|
||||
|
||||
std::vector<uint8_t> kernel_data = {
|
||||
9,
|
||||
0,
|
||||
1,
|
||||
};
|
||||
|
||||
std::vector<int32_t> bias_data = {
|
||||
-5,
|
||||
};
|
||||
|
||||
std::vector<uint8_t> golden = {
|
||||
24, 78, 27, 6, 0,
|
||||
};
|
||||
|
||||
std::vector<uint8_t> output_data(golden.size());
|
||||
|
||||
EXPECT_TRUE(
|
||||
input_tensor->CopyDataToTensor(input_data.data(), input_data.size()));
|
||||
EXPECT_TRUE(
|
||||
kernel_tensor->CopyDataToTensor(kernel_data.data(), kernel_data.size()));
|
||||
EXPECT_TRUE(bias_tensor->CopyDataToTensor(
|
||||
bias_data.data(), bias_data.size() * sizeof(int32_t)));
|
||||
|
||||
auto op = graph->CreateOperation<tim::vx::ops::DeConv1d>(
|
||||
1, tim::vx::PadType::VALID, /*ksize=*/3, /*stride=*/1,
|
||||
/*output_padding=*/0,
|
||||
std::array<uint32_t, 2>({
|
||||
0,
|
||||
0,
|
||||
}),
|
||||
/*group=*/1);
|
||||
(*op)
|
||||
.BindInputs({input_tensor, kernel_tensor, bias_tensor})
|
||||
.BindOutputs({output_tensor});
|
||||
|
||||
EXPECT_TRUE(graph->Compile());
|
||||
EXPECT_TRUE(graph->Run());
|
||||
|
||||
EXPECT_TRUE(output_tensor->CopyDataFromTensor(output_data.data()));
|
||||
EXPECT_EQ(golden, output_data) << "Result mismatch";
|
||||
}
|
||||
|
||||
TEST(DeConv1dLeakyRelu, shape_5_5_1) {
|
||||
auto ctx = tim::vx::Context::Create();
|
||||
auto graph = ctx->CreateGraph();
|
||||
|
||||
tim::vx::ShapeType input_shape({5, 5, 1}); //wcn
|
||||
tim::vx::ShapeType kernel_shape({3, 5, 2});
|
||||
tim::vx::ShapeType output_shape({7, 2, 1}); //wcn
|
||||
|
||||
tim::vx::TensorSpec input_spec(tim::vx::DataType::FLOAT32, input_shape,
|
||||
tim::vx::TensorAttribute::INPUT);
|
||||
tim::vx::TensorSpec kernel_spec(tim::vx::DataType::FLOAT32, kernel_shape,
|
||||
tim::vx::TensorAttribute::CONSTANT);
|
||||
tim::vx::TensorSpec transisent_spec(tim::vx::DataType::FLOAT32, {},
|
||||
tim::vx::TensorAttribute::TRANSIENT);
|
||||
tim::vx::TensorSpec output_spec(tim::vx::DataType::FLOAT32, output_shape,
|
||||
tim::vx::TensorAttribute::OUTPUT);
|
||||
|
||||
auto input_tensor = graph->CreateTensor(input_spec);
|
||||
auto output_tensor = graph->CreateTensor(output_spec);
|
||||
auto transient_tensor = graph->CreateTensor(transisent_spec);
|
||||
auto kernel_tensor = graph->CreateTensor(kernel_spec);
|
||||
|
||||
std::vector<float> input_data = {1.0, 2.0, 3.0, 4.0, 5.0, 6.0, 7.0,
|
||||
8.0, 9.0, 10.0, 11.0, 12.0, 13.0, 14.0,
|
||||
15.0, 16.0, 17.0, 18.0, 19.0, 20.0, 21.0,
|
||||
22.0, 23.0, 24.0, 25.0};
|
||||
std::vector<float> kernel_data = {1, 2, 3, 7, 8, 9, 13, 14, 15, 19,
|
||||
20, 21, 25, 26, 27, 4, 5, 6, 10, 11,
|
||||
12, 16, 17, 18, 22, 23, 24, 28, 29, 30};
|
||||
std::vector<float> golden = {
|
||||
1015.0, 2150.0, 3410.0, 3620.0, 3830.0, 2700.0, 1425.0,
|
||||
1180.0, 2495.0, 3950.0, 4205.0, 4460.0, 3135.0, 1650.0,
|
||||
};
|
||||
|
||||
EXPECT_TRUE(
|
||||
input_tensor->CopyDataToTensor(input_data.data(), input_data.size() * 4));
|
||||
EXPECT_TRUE(kernel_tensor->CopyDataToTensor(kernel_data.data(),
|
||||
kernel_data.size() * 4));
|
||||
|
||||
auto op = graph->CreateOperation<tim::vx::ops::DeConv1d>(
|
||||
1, tim::vx::PadType::VALID, /*ksize=*/3, /*stride=*/1,
|
||||
/*output_padding=*/0, std::array<uint32_t, 2>({0, 0}), /*group=*/1);
|
||||
(*op)
|
||||
.BindInputs({input_tensor, kernel_tensor})
|
||||
.BindOutputs({transient_tensor});
|
||||
|
||||
auto leakyrelu = graph->CreateOperation<tim::vx::ops::LeakyRelu>(0.01f);
|
||||
(*leakyrelu).BindInputs({transient_tensor}).BindOutputs({output_tensor});
|
||||
|
||||
EXPECT_TRUE(graph->Compile());
|
||||
EXPECT_TRUE(graph->Run());
|
||||
std::vector<float> output_data(golden.size());
|
||||
|
||||
EXPECT_TRUE(output_tensor->CopyDataFromTensor(output_data.data()));
|
||||
|
||||
EXPECT_EQ(golden, output_data) << "Result mismatch";
|
||||
}
|
||||
Loading…
Reference in New Issue