tensorflow/core/kernels/data/experimental/scan_dataset_op.cc - platform/external/tensorflow - Git at Google

 /* 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 <iterator>
 #include <vector>

 #include "tensorflow/core/common_runtime/function.h"
 #include "tensorflow/core/common_runtime/input_colocation_exemption_registry.h"
 #include "tensorflow/core/framework/dataset.h"
 #include "tensorflow/core/framework/partial_tensor_shape.h"
 #include "tensorflow/core/framework/tensor.h"
 #include "tensorflow/core/kernels/data/captured_function.h"
 #include "tensorflow/core/kernels/data/dataset_utils.h"
 #include "tensorflow/core/lib/random/random.h"

 namespace tensorflow {
 namespace data {
 namespace experimental {
 namespace {

 class ScanDatasetOp : public UnaryDatasetOpKernel {
  public:
   explicit ScanDatasetOp(OpKernelConstruction* ctx)
       : UnaryDatasetOpKernel(ctx) {
     FunctionMetadata::Params params;
     params.is_multi_device_function = true;
     OP_REQUIRES_OK(ctx,
                    FunctionMetadata::Create(ctx, "f", params, &func_metadata_));
     OP_REQUIRES_OK(ctx, ctx->GetAttr("Tstate", &state_types_));
     OP_REQUIRES_OK(ctx, ctx->GetAttr("output_types", &output_types_));
     OP_REQUIRES_OK(ctx, ctx->GetAttr("output_shapes", &output_shapes_));
     OP_REQUIRES_OK(
         ctx, ctx->GetAttr("preserve_cardinality", &preserve_cardinality_));
   }

   void MakeDataset(OpKernelContext* ctx, DatasetBase* input,
                    DatasetBase** output) override {
     OpInputList initial_state_inputs;
     OP_REQUIRES_OK(ctx,
                    ctx->input_list("initial_state", &initial_state_inputs));
     std::vector<Tensor> initial_state(initial_state_inputs.begin(),
                                       initial_state_inputs.end());

     std::unique_ptr<CapturedFunction> captured_func;
     OP_REQUIRES_OK(
         ctx, CapturedFunction::Create(ctx, func_metadata_, "other_arguments",
                                       &captured_func));

     *output = new Dataset(ctx, input, std::move(initial_state),
                           std::move(captured_func), state_types_, output_types_,
                           output_shapes_, preserve_cardinality_);
   }

  private:
   class Dataset : public DatasetBase {
    public:
     Dataset(OpKernelContext* ctx, const DatasetBase* input,
             std::vector<Tensor> initial_state,
             std::unique_ptr<CapturedFunction> captured_func,
             const DataTypeVector& state_types,
             const DataTypeVector& output_types,
             const std::vector<PartialTensorShape>& output_shapes,
             bool preserve_cardinality)
         : DatasetBase(DatasetContext(ctx)),
           input_(input),
           initial_state_(std::move(initial_state)),
           captured_func_(std::move(captured_func)),
           state_types_(state_types),
           output_types_(output_types),
           output_shapes_(output_shapes),
           preserve_cardinality_(preserve_cardinality) {
       input_->Ref();
     }

     ~Dataset() override { input_->Unref(); }

     std::unique_ptr<IteratorBase> MakeIteratorInternal(
         const string& prefix) const override {
       return absl::make_unique<Iterator>(
           Iterator::Params{this, strings::StrCat(prefix, "::Scan")});
     }

     const DataTypeVector& output_dtypes() const override {
       return output_types_;
     }
     const std::vector<PartialTensorShape>& output_shapes() const override {
       return output_shapes_;
     }

     string DebugString() const override { return "ScanDatasetOp::Dataset"; }

     int64 Cardinality() const override { return input_->Cardinality(); }

     Status CheckExternalState() const override {
       TF_RETURN_IF_ERROR(captured_func_->CheckExternalState());
       return input_->CheckExternalState();
     }

    protected:
     Status AsGraphDefInternal(SerializationContext* ctx,
                               DatasetGraphDefBuilder* b,
                               Node** output) const override {
       Node* input_node;
       TF_RETURN_IF_ERROR(b->AddInputDataset(ctx, input_, &input_node));
       std::vector<Node*> initial_state_nodes;
       initial_state_nodes.reserve(initial_state_.size());
       for (const Tensor& t : initial_state_) {
         Node* node;
         TF_RETURN_IF_ERROR(b->AddTensor(t, &node));
         initial_state_nodes.emplace_back(node);
       }
       std::vector<Node*> other_arguments;
       DataTypeVector other_arguments_types;
       TF_RETURN_IF_ERROR(captured_func_->AddToGraph(ctx, b, &other_arguments,
                                                     &other_arguments_types));
       AttrValue f;
       b->BuildAttrValue(captured_func_->func(), &f);
       AttrValue state_types;
       b->BuildAttrValue(state_types_, &state_types);
       AttrValue other_arguments_types_attr;
       b->BuildAttrValue(other_arguments_types, &other_arguments_types_attr);
       AttrValue preserve_cardinality_attr;
       b->BuildAttrValue(preserve_cardinality_, &preserve_cardinality_attr);
       TF_RETURN_IF_ERROR(
           b->AddDataset(this, {{0, input_node}},
                         {{1, initial_state_nodes}, {2, other_arguments}},
                         {{"f", f},
                          {"Tstate", state_types},
                          {"Targuments", other_arguments_types_attr},
                          {"preserve_cardinality", preserve_cardinality_attr}},
                         output));
       return Status::OK();
     }

    private:
     class Iterator : public DatasetIterator<Dataset> {
      public:
       explicit Iterator(const Params& params)
           : DatasetIterator<Dataset>(params),
             state_(params.dataset->initial_state_) {}

       Status Initialize(IteratorContext* ctx) override {
         TF_RETURN_IF_ERROR(
             dataset()->input_->MakeIterator(ctx, prefix(), &input_impl_));
         return dataset()->captured_func_->Instantiate(
             ctx, &instantiated_captured_func_);
       }

       Status GetNextInternal(IteratorContext* ctx,
                              std::vector<Tensor>* out_tensors,
                              bool* end_of_sequence) override {
         mutex_lock l(mu_);

         std::vector<Tensor> next_element;
         TF_RETURN_IF_ERROR(
             input_impl_->GetNext(ctx, &next_element, end_of_sequence));
         if (*end_of_sequence) {
           return Status::OK();
         }

         std::vector<Tensor> args;
         args.reserve(state_.size() + next_element.size());
         std::copy(state_.begin(), state_.end(), std::back_inserter(args));
         std::copy(next_element.begin(), next_element.end(),
                   std::back_inserter(args));

         std::vector<Tensor> state_and_output;
         state_and_output.reserve(dataset()->state_types_.size() +
                                  output_dtypes().size());

         Status s = instantiated_captured_func_->Run(ctx, std::move(args),
                                                     &state_and_output);
         DCHECK(state_and_output.size() <=
                dataset()->state_types_.size() + output_dtypes().size());
         if (s.ok()) {
           state_.clear();
           size_t i = 0;
           for (; i < dataset()->state_types_.size(); ++i) {
             if (state_and_output[i].dtype() != dataset()->state_types_[i]) {
               return errors::InvalidArgument(
                   "Got wrong type for scan_func return value ", i,
                   " (expected ", DataTypeString(dataset()->state_types_[i]),
                   ", got ", DataTypeString(state_and_output[i].dtype()), ").");
             }
             state_.push_back(std::move(state_and_output[i]));
           }
           for (; i < state_and_output.size(); ++i) {
             const size_t output_index = i - dataset()->state_types_.size();
             if (state_and_output[i].dtype() != output_dtypes()[output_index]) {
               return errors::InvalidArgument(
                   "Got wrong type for scan_func return value ", i,
                   " (expected ",
                   DataTypeString(dataset()->state_types_[output_index]),
                   ", got ", DataTypeString(state_and_output[i].dtype()), ").");
             }
             if (!output_shapes()[output_index].IsCompatibleWith(
                     state_and_output[i].shape())) {
               return errors::InvalidArgument(
                   "Got wrong shape for scan_func return value ", i,
                   " (expected ", output_shapes()[output_index].DebugString(),
                   ", got ", state_and_output[i].shape().DebugString(), ").");
             }

             out_tensors->push_back(std::move(state_and_output[i]));
           }
         } else if (errors::IsOutOfRange(s)) {
           if (dataset()->preserve_cardinality_) {
             // To guarantee that the transformation preserves the cardinality of
             // the dataset, we convert `OutOfRange` to `InvalidArgument` as the
             // former may be interpreted by a caller as the end of sequence.
             return errors::InvalidArgument(
                 "Function invocation produced OutOfRangeError: ",
                 s.error_message());
           } else {
             // `f` may deliberately raise `errors::OutOfRange` to indicate
             // that we should terminate the iteration early.
             *end_of_sequence = true;
             return Status::OK();
           }
         }
         return s;
       }

      protected:
       std::shared_ptr<model::Node> CreateNode(
           IteratorContext* ctx, model::Node::Args args) const override {
         return model::MakeKnownRatioNode(std::move(args),
                                          /*ratio=*/1);
       }

       Status SaveInternal(IteratorStateWriter* writer) override {
         mutex_lock l(mu_);
         TF_RETURN_IF_ERROR(SaveInput(writer, input_impl_));
         if (!state_.empty()) {
           TF_RETURN_IF_ERROR(
               writer->WriteScalar(full_name("state_size"), state_.size()));
           for (int idx = 0; idx < state_.size(); idx++) {
             TF_RETURN_IF_ERROR(writer->WriteTensor(
                 full_name(strings::StrCat("state[", idx, "]")), state_[idx]));
           }
         }
         return Status::OK();
       }

       Status RestoreInternal(IteratorContext* ctx,
                              IteratorStateReader* reader) override {
         mutex_lock l(mu_);
         TF_RETURN_IF_ERROR(RestoreInput(ctx, reader, input_impl_));
         if (reader->Contains(full_name("state_size"))) {
           int64 size;
           TF_RETURN_IF_ERROR(
               reader->ReadScalar(full_name("state_size"), &size));
           state_.resize(size);
           for (int idx = 0; idx < size; idx++) {
             TF_RETURN_IF_ERROR(reader->ReadTensor(
                 full_name(strings::StrCat("state[", idx, "]")), &state_[idx]));
           }
         }
         return Status::OK();
       }

      private:
       mutex mu_;
       std::unique_ptr<IteratorBase> input_impl_ GUARDED_BY(mu_);
       std::vector<Tensor> state_ GUARDED_BY(mu_);
       std::unique_ptr<InstantiatedCapturedFunction> instantiated_captured_func_;
     };

     const DatasetBase* const input_;
     const std::vector<Tensor> initial_state_;
     const std::unique_ptr<CapturedFunction> captured_func_;
     const DataTypeVector state_types_;
     const DataTypeVector output_types_;
     const std::vector<PartialTensorShape> output_shapes_;
     const bool preserve_cardinality_;
   };

   std::shared_ptr<FunctionMetadata> func_metadata_ = nullptr;
   DataTypeVector state_types_;
   DataTypeVector output_types_;
   std::vector<PartialTensorShape> output_shapes_;
   bool preserve_cardinality_;
 };

 REGISTER_KERNEL_BUILDER(Name("ScanDataset").Device(DEVICE_CPU), ScanDatasetOp);
 REGISTER_KERNEL_BUILDER(Name("ExperimentalScanDataset").Device(DEVICE_CPU),
                         ScanDatasetOp);

 REGISTER_INPUT_COLOCATION_EXEMPTION("ScanDataset");
 REGISTER_INPUT_COLOCATION_EXEMPTION("ExperimentalScanDataset");

 }  // namespace
 }  // namespace experimental
 }  // namespace data
 }  // namespace tensorflow
	/* 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 <iterator>
	#include <vector>

	#include "tensorflow/core/common_runtime/function.h"
	#include "tensorflow/core/common_runtime/input_colocation_exemption_registry.h"
	#include "tensorflow/core/framework/dataset.h"
	#include "tensorflow/core/framework/partial_tensor_shape.h"
	#include "tensorflow/core/framework/tensor.h"
	#include "tensorflow/core/kernels/data/captured_function.h"
	#include "tensorflow/core/kernels/data/dataset_utils.h"
	#include "tensorflow/core/lib/random/random.h"

	namespace tensorflow {
	namespace data {
	namespace experimental {
	namespace {

	class ScanDatasetOp : public UnaryDatasetOpKernel {
	public:
	explicit ScanDatasetOp(OpKernelConstruction* ctx)
	: UnaryDatasetOpKernel(ctx) {
	FunctionMetadata::Params params;
	params.is_multi_device_function = true;
	OP_REQUIRES_OK(ctx,
	FunctionMetadata::Create(ctx, "f", params, &func_metadata_));
	OP_REQUIRES_OK(ctx, ctx->GetAttr("Tstate", &state_types_));
	OP_REQUIRES_OK(ctx, ctx->GetAttr("output_types", &output_types_));
	OP_REQUIRES_OK(ctx, ctx->GetAttr("output_shapes", &output_shapes_));
	OP_REQUIRES_OK(
	ctx, ctx->GetAttr("preserve_cardinality", &preserve_cardinality_));
	}

	void MakeDataset(OpKernelContext* ctx, DatasetBase* input,
	DatasetBase** output) override {
	OpInputList initial_state_inputs;
	OP_REQUIRES_OK(ctx,
	ctx->input_list("initial_state", &initial_state_inputs));
	std::vector<Tensor> initial_state(initial_state_inputs.begin(),
	initial_state_inputs.end());

	std::unique_ptr<CapturedFunction> captured_func;
	OP_REQUIRES_OK(
	ctx, CapturedFunction::Create(ctx, func_metadata_, "other_arguments",
	&captured_func));

	*output = new Dataset(ctx, input, std::move(initial_state),
	std::move(captured_func), state_types_, output_types_,
	output_shapes_, preserve_cardinality_);
	}

	private:
	class Dataset : public DatasetBase {
	public:
	Dataset(OpKernelContext* ctx, const DatasetBase* input,
	std::vector<Tensor> initial_state,
	std::unique_ptr<CapturedFunction> captured_func,
	const DataTypeVector& state_types,
	const DataTypeVector& output_types,
	const std::vector<PartialTensorShape>& output_shapes,
	bool preserve_cardinality)
	: DatasetBase(DatasetContext(ctx)),
	input_(input),
	initial_state_(std::move(initial_state)),
	captured_func_(std::move(captured_func)),
	state_types_(state_types),
	output_types_(output_types),
	output_shapes_(output_shapes),
	preserve_cardinality_(preserve_cardinality) {
	input_->Ref();
	}

	~Dataset() override { input_->Unref(); }

	std::unique_ptr<IteratorBase> MakeIteratorInternal(
	const string& prefix) const override {
	return absl::make_unique<Iterator>(
	Iterator::Params{this, strings::StrCat(prefix, "::Scan")});
	}

	const DataTypeVector& output_dtypes() const override {
	return output_types_;
	}
	const std::vector<PartialTensorShape>& output_shapes() const override {
	return output_shapes_;
	}

	string DebugString() const override { return "ScanDatasetOp::Dataset"; }

	int64 Cardinality() const override { return input_->Cardinality(); }

	Status CheckExternalState() const override {
	TF_RETURN_IF_ERROR(captured_func_->CheckExternalState());
	return input_->CheckExternalState();
	}

	protected:
	Status AsGraphDefInternal(SerializationContext* ctx,
	DatasetGraphDefBuilder* b,
	Node** output) const override {
	Node* input_node;
	TF_RETURN_IF_ERROR(b->AddInputDataset(ctx, input_, &input_node));
	std::vector<Node*> initial_state_nodes;
	initial_state_nodes.reserve(initial_state_.size());
	for (const Tensor& t : initial_state_) {
	Node* node;
	TF_RETURN_IF_ERROR(b->AddTensor(t, &node));
	initial_state_nodes.emplace_back(node);
	}
	std::vector<Node*> other_arguments;
	DataTypeVector other_arguments_types;
	TF_RETURN_IF_ERROR(captured_func_->AddToGraph(ctx, b, &other_arguments,
	&other_arguments_types));
	AttrValue f;
	b->BuildAttrValue(captured_func_->func(), &f);
	AttrValue state_types;
	b->BuildAttrValue(state_types_, &state_types);
	AttrValue other_arguments_types_attr;
	b->BuildAttrValue(other_arguments_types, &other_arguments_types_attr);
	AttrValue preserve_cardinality_attr;
	b->BuildAttrValue(preserve_cardinality_, &preserve_cardinality_attr);
	TF_RETURN_IF_ERROR(
	b->AddDataset(this, {{0, input_node}},
	{{1, initial_state_nodes}, {2, other_arguments}},
	{{"f", f},
	{"Tstate", state_types},
	{"Targuments", other_arguments_types_attr},
	{"preserve_cardinality", preserve_cardinality_attr}},
	output));
	return Status::OK();
	}

	private:
	class Iterator : public DatasetIterator<Dataset> {
	public:
	explicit Iterator(const Params& params)
	: DatasetIterator<Dataset>(params),
	state_(params.dataset->initial_state_) {}

	Status Initialize(IteratorContext* ctx) override {
	TF_RETURN_IF_ERROR(
	dataset()->input_->MakeIterator(ctx, prefix(), &input_impl_));
	return dataset()->captured_func_->Instantiate(
	ctx, &instantiated_captured_func_);
	}

	Status GetNextInternal(IteratorContext* ctx,
	std::vector<Tensor>* out_tensors,
	bool* end_of_sequence) override {
	mutex_lock l(mu_);

	std::vector<Tensor> next_element;
	TF_RETURN_IF_ERROR(
	input_impl_->GetNext(ctx, &next_element, end_of_sequence));
	if (*end_of_sequence) {
	return Status::OK();
	}

	std::vector<Tensor> args;
	args.reserve(state_.size() + next_element.size());
	std::copy(state_.begin(), state_.end(), std::back_inserter(args));
	std::copy(next_element.begin(), next_element.end(),
	std::back_inserter(args));

	std::vector<Tensor> state_and_output;
	state_and_output.reserve(dataset()->state_types_.size() +
	output_dtypes().size());

	Status s = instantiated_captured_func_->Run(ctx, std::move(args),
	&state_and_output);
	DCHECK(state_and_output.size() <=
	dataset()->state_types_.size() + output_dtypes().size());
	if (s.ok()) {
	state_.clear();
	size_t i = 0;
	for (; i < dataset()->state_types_.size(); ++i) {
	if (state_and_output[i].dtype() != dataset()->state_types_[i]) {
	return errors::InvalidArgument(
	"Got wrong type for scan_func return value ", i,
	" (expected ", DataTypeString(dataset()->state_types_[i]),
	", got ", DataTypeString(state_and_output[i].dtype()), ").");
	}
	state_.push_back(std::move(state_and_output[i]));
	}
	for (; i < state_and_output.size(); ++i) {
	const size_t output_index = i - dataset()->state_types_.size();
	if (state_and_output[i].dtype() != output_dtypes()[output_index]) {
	return errors::InvalidArgument(
	"Got wrong type for scan_func return value ", i,
	" (expected ",
	DataTypeString(dataset()->state_types_[output_index]),
	", got ", DataTypeString(state_and_output[i].dtype()), ").");
	}
	if (!output_shapes()[output_index].IsCompatibleWith(
	state_and_output[i].shape())) {
	return errors::InvalidArgument(
	"Got wrong shape for scan_func return value ", i,
	" (expected ", output_shapes()[output_index].DebugString(),
	", got ", state_and_output[i].shape().DebugString(), ").");
	}

	out_tensors->push_back(std::move(state_and_output[i]));
	}
	} else if (errors::IsOutOfRange(s)) {
	if (dataset()->preserve_cardinality_) {
	// To guarantee that the transformation preserves the cardinality of
	// the dataset, we convert `OutOfRange` to `InvalidArgument` as the
	// former may be interpreted by a caller as the end of sequence.
	return errors::InvalidArgument(
	"Function invocation produced OutOfRangeError: ",
	s.error_message());
	} else {
	// `f` may deliberately raise `errors::OutOfRange` to indicate
	// that we should terminate the iteration early.
	*end_of_sequence = true;
	return Status::OK();
	}
	}
	return s;
	}

	protected:
	std::shared_ptr<model::Node> CreateNode(
	IteratorContext* ctx, model::Node::Args args) const override {
	return model::MakeKnownRatioNode(std::move(args),
	/ratio=/1);
	}

	Status SaveInternal(IteratorStateWriter* writer) override {
	mutex_lock l(mu_);
	TF_RETURN_IF_ERROR(SaveInput(writer, input_impl_));
	if (!state_.empty()) {
	TF_RETURN_IF_ERROR(
	writer->WriteScalar(full_name("state_size"), state_.size()));
	for (int idx = 0; idx < state_.size(); idx++) {
	TF_RETURN_IF_ERROR(writer->WriteTensor(
	full_name(strings::StrCat("state[", idx, "]")), state_[idx]));
	}
	}
	return Status::OK();
	}

	Status RestoreInternal(IteratorContext* ctx,
	IteratorStateReader* reader) override {
	mutex_lock l(mu_);
	TF_RETURN_IF_ERROR(RestoreInput(ctx, reader, input_impl_));
	if (reader->Contains(full_name("state_size"))) {
	int64 size;
	TF_RETURN_IF_ERROR(
	reader->ReadScalar(full_name("state_size"), &size));
	state_.resize(size);
	for (int idx = 0; idx < size; idx++) {
	TF_RETURN_IF_ERROR(reader->ReadTensor(
	full_name(strings::StrCat("state[", idx, "]")), &state_[idx]));
	}
	}
	return Status::OK();
	}

	private:
	mutex mu_;
	std::unique_ptr<IteratorBase> input_impl_ GUARDED_BY(mu_);
	std::vector<Tensor> state_ GUARDED_BY(mu_);
	std::unique_ptr<InstantiatedCapturedFunction> instantiated_captured_func_;
	};

	const DatasetBase* const input_;
	const std::vector<Tensor> initial_state_;
	const std::unique_ptr<CapturedFunction> captured_func_;
	const DataTypeVector state_types_;
	const DataTypeVector output_types_;
	const std::vector<PartialTensorShape> output_shapes_;
	const bool preserve_cardinality_;
	};

	std::shared_ptr<FunctionMetadata> func_metadata_ = nullptr;
	DataTypeVector state_types_;
	DataTypeVector output_types_;
	std::vector<PartialTensorShape> output_shapes_;
	bool preserve_cardinality_;
	};

	REGISTER_KERNEL_BUILDER(Name("ScanDataset").Device(DEVICE_CPU), ScanDatasetOp);
	REGISTER_KERNEL_BUILDER(Name("ExperimentalScanDataset").Device(DEVICE_CPU),
	ScanDatasetOp);

	REGISTER_INPUT_COLOCATION_EXEMPTION("ScanDataset");
	REGISTER_INPUT_COLOCATION_EXEMPTION("ExperimentalScanDataset");

	} // namespace
	} // namespace experimental
	} // namespace data
	} // namespace tensorflow