experiments/STT-tensorflow
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

Fix some old TODO comments: delete methods of Subgraph that expose

Browse Source 
mutable access to the private `tensors_` vector and to the private
`nodes_and_registrations_` vector, replacing them with functions that
provide mutable access to the individual tensors/nodes/registrations,
and apply minor refactorings to the code that was previously using these
methods to use other methods instead.

PiperOrigin-RevId: 345649125
Change-Id: I1b38c22421d62ef85c7a1eadf9662234995c90fe
This commit is contained in:
Fergus Henderson 2020-12-04 04:07:24 -08:00 committed by TensorFlower Gardener
parent 0fd465c057
commit c37b2f1152
5 changed files with 55 additions and 52 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

10
RELEASE.md
View File

@ -42,6 +42,16 @@
renamed `tf.function(jit_compile=True)`.
* `tf.lite`:
* class `tflite::Subgraph`:
* Removed the `tensors()` method and the non-const overload of the
`nodes_and_registration()` method, both of which were previously
documented as temporary and to be removed.
* Uses of `tensors()` can be replaced by calling the existing
methods `tensors_size()` and `tensor(int)`.
* Uses of the non-const overload of `nodes_and_registration`
can be replaced by calling the existing methods `nodes_size()`
and `context()`, and then calling the `GetNodeAndRegistration`
method in the `TfLiteContext` returned by `context()`.
* NNAPI
* Removed deprecated `Interpreter::UseNNAPI(bool)` C++ API.
* Use `NnApiDelegate()` and related delegate configuration methods

17
tensorflow/lite/core/subgraph.cc
View File

@ -162,9 +162,9 @@ class InterpreterInfo : public GraphInfo {
public:
explicit InterpreterInfo(Subgraph* subgraph) : subgraph_(subgraph) {}
size_t num_tensors() const override { return subgraph_->tensors().size(); }
size_t num_tensors() const override { return subgraph_->tensors_size(); }
TfLiteTensor* tensor(size_t index) override {
return &subgraph_->tensors()[index];
return subgraph_->tensor(index);
}
size_t num_execution_nodes() const override {
return subgraph_->execution_plan().size();
@ -218,7 +218,7 @@ Subgraph::Subgraph(ErrorReporter* error_reporter,
// Reserve some space for the tensors to avoid excessive resizing.
tensors_.reserve(kTensorsReservedCapacity);
nodes_and_registration().reserve(kTensorsReservedCapacity);
nodes_and_registration_.reserve(kTensorsReservedCapacity);
// Invalid to call these except from TfLiteDelegate
SwitchToKernelContext();
}
@ -961,12 +961,13 @@ TfLiteStatus Subgraph::PrepareOpsAndTensors() {
// overhead should be minimal since the number of custom-allocated tensors
// will typically be low.
for (int i = 0; i < custom_allocations_.size(); ++i) {
auto idx_and_alloc = custom_allocations_[i];
auto& tensor = tensors()[idx_and_alloc.first];
const auto& alloc = idx_and_alloc.second;
TF_LITE_ENSURE(context(), tensor.allocation_type == kTfLiteCustom);
auto index_and_alloc = custom_allocations_[i];
TfLiteTensor* tensor_at_index = tensor(index_and_alloc.first);
const auto& alloc = index_and_alloc.second;
TF_LITE_ENSURE(context(),
tensor_at_index->allocation_type == kTfLiteCustom);
TF_LITE_ENSURE_STATUS(
ValidateCustomAllocationForTensor(context(), &tensor, alloc));
ValidateCustomAllocationForTensor(context(), tensor_at_index, alloc));
}
next_execution_plan_index_to_plan_allocation_ =

11
tensorflow/lite/core/subgraph.h
View File

@ -180,17 +180,6 @@ class Subgraph {
// Return read-only vector of node indices in the order of execution.
const std::vector<int>& execution_plan() const { return execution_plan_; }
// Mutable form of tensors (TEMPORARY for refactor).
// TODO(b/119495520): remove when refactoring complete.
std::vector<TfLiteTensor>& tensors() { return tensors_; }
// Mutable form of nodes_and_registration (TEMPORARY for refactor).
// TODO(b/119495520): remove when refactoring complete.
std::vector<std::pair<TfLiteNode, TfLiteRegistration>>&
nodes_and_registration() {
return nodes_and_registration_;
}
const std::vector<std::pair<TfLiteNode, TfLiteRegistration>>&
nodes_and_registration() const {
return nodes_and_registration_;

63
tensorflow/lite/delegates/gpu/common/model_builder_test.cc
View File

@ -141,12 +141,28 @@ class DelegatedInterpreter {
}
// Get the TfLiteContext to be mocked for swapping out functions that have to
// be called inside delegate (i.e. in delegat kernel mode).
// be called inside delegate (i.e. in delegate kernel mode).
TfLiteContext* context() { return interpreter_.primary_subgraph().context(); }
std::vector<std::pair<TfLiteNode, TfLiteRegistration>>&
nodes_and_registration() {
return interpreter_.primary_subgraph().nodes_and_registration();
// node(int) and registration(int) are used to implement
// GetNodeAndRegistration. We can't implement those using
// TfLiteContext *context = interpreter_.primary_subgraph().context();
// context->GetNodeAndRegistration(context, &node, &registration);
// here, because calling GetNodeAndRegistration from within it's own
// implementation would lead to an infinite loop.
// Instead, we just call node_and_registration and use a const_cast.
// These const_casts are a bit ugly, but I think less ugly than exposing
// the private GetNodeAndRegistration method in Subgraph as public,
// or making this class a friend of Subgraph.
TfLiteNode* node(int index) {
const std::pair<TfLiteNode, TfLiteRegistration>* node_and_registration =
interpreter_.primary_subgraph().node_and_registration(index);
return const_cast<TfLiteNode*>(&node_and_registration->first);
}
TfLiteRegistration* registration(int index) {
const std::pair<TfLiteNode, TfLiteRegistration>* node_and_registration =
interpreter_.primary_subgraph().node_and_registration(index);
return const_cast<TfLiteRegistration*>(&node_and_registration->second);
}
TfLiteIntArray* exec_plan() const { return exec_plan_; }
@ -280,10 +296,8 @@ TEST(ModelBuilderTest, GetOpsToReplaceAcceptsFp16DequantizeNodes) {
context->GetNodeAndRegistration = [](struct TfLiteContext*, int node_index,
TfLiteNode** node,
TfLiteRegistration** registration) {
auto& node_and_reg =
interpreter_fp16_add_op->nodes_and_registration()[node_index];
*node = &node_and_reg.first;
*registration = &node_and_reg.second;
*node = interpreter_fp16_add_op->node(node_index);
*registration = interpreter_fp16_add_op->registration(node_index);
return kTfLiteOk;
};
context->PreviewDelegatePartitioning =
@ -346,10 +360,8 @@ TEST(ModelBuilderTest, GetOpsToReplaceRejectsFp16DequantizeNodes) {
context->GetNodeAndRegistration = [](struct TfLiteContext*, int node_index,
TfLiteNode** node,
TfLiteRegistration** registration) {
auto& node_and_reg =
interpreter_fp16_gt_op->nodes_and_registration()[node_index];
*node = &node_and_reg.first;
*registration = &node_and_reg.second;
*node = interpreter_fp16_gt_op->node(node_index);
*registration = interpreter_fp16_gt_op->registration(node_index);
return kTfLiteOk;
};
context->PreviewDelegatePartitioning =
@ -462,9 +474,8 @@ TEST(ModelBuilderTest, GetOpsToReplaceDoesNotPruneUint8) {
context->GetNodeAndRegistration = [](struct TfLiteContext*, int node_index,
TfLiteNode** node,
TfLiteRegistration** registration) {
auto& node_and_reg = interpreter_fp32->nodes_and_registration()[node_index];
*node = &node_and_reg.first;
*registration = &node_and_reg.second;
*node = interpreter_fp32->node(node_index);
*registration = interpreter_fp32->registration(node_index);
return kTfLiteOk;
};
context->PreviewDelegatePartitioning =
@ -630,10 +641,8 @@ TEST(ModelBuilderTest, GetOpsToReplaceMultiplePartitions) {
context->GetNodeAndRegistration = [](struct TfLiteContext*, int node_index,
TfLiteNode** node,
TfLiteRegistration** registration) {
auto& node_and_reg =
interpreter2_fp32->nodes_and_registration()[node_index];
*node = &node_and_reg.first;
*registration = &node_and_reg.second;
*node = interpreter2_fp32->node(node_index);
*registration = interpreter2_fp32->registration(node_index);
return kTfLiteOk;
};
context->PreviewDelegatePartitioning =
@ -845,9 +854,8 @@ TEST(ModelBuilderTest, GetOpsToReplaceSelectsCorrectFp16Nodes_SinglePartition) {
context->GetNodeAndRegistration = [](struct TfLiteContext*, int node_index,
TfLiteNode** node,
TfLiteRegistration** registration) {
auto& node_and_reg = interpreter_mn->nodes_and_registration()[node_index];
*node = &node_and_reg.first;
*registration = &node_and_reg.second;
*node = interpreter_mn->node(node_index);
*registration = interpreter_mn->registration(node_index);
return kTfLiteOk;
};
context->PreviewDelegatePartitioning =
@ -917,9 +925,8 @@ TEST(ModelBuilderTest,
context->GetNodeAndRegistration = [](struct TfLiteContext*, int node_index,
TfLiteNode** node,
TfLiteRegistration** registration) {
auto& node_and_reg = interpreter_mn2->nodes_and_registration()[node_index];
*node = &node_and_reg.first;
*registration = &node_and_reg.second;
*node = interpreter_mn2->node(node_index);
*registration = interpreter_mn2->registration(node_index);
return kTfLiteOk;
};
@ -1121,10 +1128,8 @@ TEST(ModelBuilderTest, GetOpsToReplace_AllowQuantOps) {
context->GetNodeAndRegistration = [](struct TfLiteContext*, int node_index,
TfLiteNode** node,
TfLiteRegistration** registration) {
auto& node_and_reg =
interpreter_quant->nodes_and_registration()[node_index];
*node = &node_and_reg.first;
*registration = &node_and_reg.second;
*node = interpreter_quant->node(node_index);
*registration = interpreter_quant->registration(node_index);
return kTfLiteOk;
};
context->PreviewDelegatePartitioning =

6
tensorflow/lite/interpreter.cc
View File

@ -411,8 +411,7 @@ TfLiteStatus Interpreter::SetBufferHandle(int tensor_index,
TfLiteBufferHandle buffer_handle,
TfLiteDelegate* delegate) {
TF_LITE_ENSURE(context_, tensor_index < tensors_size());
std::vector<TfLiteTensor>& tensors = primary_subgraph().tensors();
TfLiteTensor* tensor = &tensors[tensor_index];
TfLiteTensor* tensor = primary_subgraph().tensor(tensor_index);
TF_LITE_ENSURE(context_,
tensor->delegate == nullptr || tensor->delegate == delegate);
@ -431,8 +430,7 @@ TfLiteStatus Interpreter::GetBufferHandle(int tensor_index,
TfLiteBufferHandle* buffer_handle,
TfLiteDelegate** delegate) {
TF_LITE_ENSURE(context_, tensor_index < tensors_size());
std::vector<TfLiteTensor>& tensors = primary_subgraph().tensors();
TfLiteTensor* tensor = &tensors[tensor_index];
TfLiteTensor* tensor = primary_subgraph().tensor(tensor_index);
*delegate = tensor->delegate;
*buffer_handle = tensor->buffer_handle;