wehub-resource-sync
132 lines
4.8 KiB
C++
132 lines
4.8 KiB
C++
/* Copyright 2017 The TensorFlow Authors. All Rights Reserved.
|
|
|
|
Licensed under the Apache License, Version 2.0 (the "License");
|
|
you may not use this file except in compliance with the License.
|
|
You may obtain a copy of the License at
|
|
|
|
http://www.apache.org/licenses/LICENSE-2.0
|
|
|
|
Unless required by applicable law or agreed to in writing, software
|
|
distributed under the License is distributed on an "AS IS" BASIS,
|
|
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
|
See the License for the specific language governing permissions and
|
|
limitations under the License.
|
|
==============================================================================*/
|
|
#include "tensorflow/lite/optional_debug_tools.h"
|
|
|
|
#include <algorithm>
|
|
#include <memory>
|
|
|
|
#include <gtest/gtest.h>
|
|
#include "tensorflow/lite/core/interpreter.h"
|
|
#include "tensorflow/lite/core/interpreter_builder.h"
|
|
#include "tensorflow/lite/core/kernels/register.h"
|
|
#include "tensorflow/lite/core/model_builder.h"
|
|
#include "tensorflow/lite/delegates/xnnpack/xnnpack_delegate.h"
|
|
|
|
namespace tflite {
|
|
namespace {
|
|
// This is specific to the testdata/add.bin model used in the test.
|
|
void InitInputTensorData(Interpreter* interpreter) {
|
|
ASSERT_EQ(interpreter->inputs().size(), 1);
|
|
TfLiteTensor* t = interpreter->input_tensor(0);
|
|
ASSERT_EQ(t->type, kTfLiteFloat32);
|
|
float* data = static_cast<float*>(t->data.data);
|
|
int num_elements = t->bytes / sizeof(float);
|
|
std::fill(data, data + num_elements, 1.0f);
|
|
}
|
|
} // namespace
|
|
|
|
TEST(OptionalDebugTools, PrintInterpreterState) {
|
|
auto model = FlatBufferModel::BuildFromFile(
|
|
"tensorflow/lite/testdata/add.bin");
|
|
ASSERT_TRUE(model);
|
|
|
|
std::unique_ptr<Interpreter> interpreter;
|
|
ASSERT_EQ(
|
|
InterpreterBuilder(
|
|
*model, ops::builtin::BuiltinOpResolverWithoutDefaultDelegates())(
|
|
&interpreter),
|
|
kTfLiteOk);
|
|
|
|
// Ensure printing the interpreter state doesn't crash:
|
|
// * Before allocation
|
|
// * After allocation
|
|
// * After invocation
|
|
PrintInterpreterState(interpreter.get());
|
|
|
|
ASSERT_EQ(interpreter->AllocateTensors(), kTfLiteOk);
|
|
PrintInterpreterState(interpreter.get());
|
|
|
|
InitInputTensorData(interpreter.get());
|
|
ASSERT_EQ(interpreter->Invoke(), kTfLiteOk);
|
|
PrintInterpreterState(interpreter.get());
|
|
}
|
|
|
|
TEST(OptionalDebugTools, PrintInterpreterStateWithDelegate) {
|
|
auto model = FlatBufferModel::BuildFromFile(
|
|
"tensorflow/lite/testdata/add.bin");
|
|
ASSERT_TRUE(model);
|
|
|
|
// Create and instantiate an interpreter with a delegate.
|
|
std::unique_ptr<TfLiteDelegate, decltype(&TfLiteXNNPackDelegateDelete)>
|
|
xnnpack_delegate(TfLiteXNNPackDelegateCreate(nullptr),
|
|
TfLiteXNNPackDelegateDelete);
|
|
std::unique_ptr<Interpreter> interpreter;
|
|
ASSERT_EQ(
|
|
InterpreterBuilder(
|
|
*model, ops::builtin::BuiltinOpResolverWithoutDefaultDelegates())(
|
|
&interpreter),
|
|
kTfLiteOk);
|
|
ASSERT_EQ(interpreter->AllocateTensors(), kTfLiteOk);
|
|
ASSERT_EQ(interpreter->ModifyGraphWithDelegate(xnnpack_delegate.get()),
|
|
kTfLiteOk);
|
|
|
|
InitInputTensorData(interpreter.get());
|
|
ASSERT_EQ(interpreter->Invoke(), kTfLiteOk);
|
|
|
|
// Ensure printing the interpreter state doesn't crash.
|
|
PrintInterpreterState(interpreter.get());
|
|
}
|
|
|
|
TEST(OptionalDebugTools, GetNodeDelegationMetadata) {
|
|
auto model = FlatBufferModel::BuildFromFile(
|
|
"tensorflow/lite/testdata/add.bin");
|
|
ASSERT_TRUE(model);
|
|
|
|
// Create and instantiate an interpreter with a delegate.
|
|
std::unique_ptr<Interpreter> interpreter;
|
|
ASSERT_EQ(
|
|
InterpreterBuilder(
|
|
*model, ops::builtin::BuiltinOpResolverWithoutDefaultDelegates())(
|
|
&interpreter),
|
|
kTfLiteOk);
|
|
|
|
ASSERT_EQ(interpreter->AllocateTensors(), kTfLiteOk);
|
|
|
|
auto metadata = GetNodeDelegationMetadata(*interpreter->subgraph(0));
|
|
EXPECT_FALSE(metadata.has_delegate_applied);
|
|
for (int i = 0; i < metadata.is_node_delegated.size(); ++i) {
|
|
EXPECT_FALSE(metadata.is_node_delegated[i]);
|
|
EXPECT_EQ(metadata.replaced_by_node[i], -1);
|
|
}
|
|
|
|
std::unique_ptr<TfLiteDelegate, decltype(&TfLiteXNNPackDelegateDelete)>
|
|
xnnpack_delegate(TfLiteXNNPackDelegateCreate(nullptr),
|
|
TfLiteXNNPackDelegateDelete);
|
|
ASSERT_EQ(interpreter->ModifyGraphWithDelegate(xnnpack_delegate.get()),
|
|
kTfLiteOk);
|
|
auto metadata_with_delegate =
|
|
GetNodeDelegationMetadata(*interpreter->subgraph(0));
|
|
|
|
// The first two nodes are delegated to the third node.
|
|
EXPECT_TRUE(metadata_with_delegate.has_delegate_applied);
|
|
EXPECT_EQ(metadata_with_delegate.is_node_delegated[0], true);
|
|
EXPECT_EQ(metadata_with_delegate.replaced_by_node[0], 2);
|
|
EXPECT_EQ(metadata_with_delegate.is_node_delegated[1], true);
|
|
EXPECT_EQ(metadata_with_delegate.replaced_by_node[1], 2);
|
|
EXPECT_EQ(metadata_with_delegate.is_node_delegated[2], false);
|
|
EXPECT_EQ(metadata_with_delegate.replaced_by_node[2], -1);
|
|
}
|
|
} // namespace tflite
|