namespace tensorflow {
namespace parallel_device {

TEST(PARALLEL_DEVICE, TestRemoteBasic) {
  std::unique_ptr<TFE_ContextOptions, decltype(&TFE_DeleteContextOptions)> opts(
      TFE_NewContextOptions(), TFE_DeleteContextOptions);
  std::unique_ptr<TF_Status, decltype(&TF_DeleteStatus)> status(
      TF_NewStatus(), TF_DeleteStatus);
  std::unique_ptr<TFE_Context, decltype(&TFE_DeleteContext)> context(

    // Input tensors
    const std::vector<std::pair<string, Tensor>>& input_pairs,
    // Output tensors
    const std::vector<string>& output_tensor_names, TF_Tensor** c_outputs,
    // Target nodes
    const std::vector<string>& target_oper_names, TF_Buffer* run_metadata,
    TF_Status* status) {
  const int noutputs = output_tensor_names.size();
  std::vector<Tensor> outputs(noutputs);
  Status result;

    DCHECK(upstream_grad);

    // Recover data format from forward pass for gradient.
    std::string data_format;
    TF_RETURN_IF_ERROR(forward_attrs_.Get("data_format", &data_format));

    // Grad for A
    grad_inputs[0] = upstream_grad;
    grad_inputs[0]->Ref();

    // Grad for bias
    std::string name = "bias_add_grad";
    TF_RETURN_IF_ERROR(BiasAddGrad(ctx, upstream_grad, &grad_inputs[1],

namespace {

void TestHandleFromDLPack(TF_Status* status, TFE_Context* ctx,
                          std::vector<int64_t> shape,
                          std::vector<int64_t> strides) {
  size_t num_elements = 1;
  for (int i = 0; i < static_cast<int32_t>(shape.size()); ++i) {
    num_elements *= shape[i];
  }
  std::vector<float> data(num_elements);
  for (size_t j = 0; j < num_elements; ++j) {
    data[j] = j;
  }

#include "tensorflow/core/platform/errors.h"
#include "tensorflow/core/platform/types.h"

using tensorflow::string;

namespace tensorflow {
namespace tracing {
typedef absl::flat_hash_map<std::string, tracing::FactoryFunction> FactoriesMap;

static FactoriesMap& GetFactories() {
  static FactoriesMap* factories = new FactoriesMap;
  return *factories;
}

static tracing::FactoryFunction default_factory;

namespace tensorflow {
namespace gradients {
namespace internal {
namespace {
using std::vector;

class CustomGradientTest
    : public ::testing::TestWithParam<std::tuple<const char*, bool, bool>> {
 protected:
  void SetUp() override {
    TF_StatusPtr status(TF_NewStatus());
    TF_SetTracingImplementation(std::get<0>(GetParam()), status.get());
    Status s = StatusFromTF_Status(status.get());

  InferenceContext c(TF_GRAPH_DEF_VERSION, node_def, op_reg_data->op_def,
                     std::vector<ShapeHandle>(num_inputs), input_tensors_vector,
                     {},
                     std::vector<std::unique_ptr<std::vector<ShapeAndType>>>());

  // Set input_shapes.
  for (int i = 0; i < num_inputs; ++i) {
    std::vector<DimensionHandle> dims;
    const TF_ShapeAndType& input_shape = input_shapes->items[i];

    }
  }

  TF_DeleteTensor(analytical_tensor);
  delete[] danalytical;
}

Model BuildGradModel(Model forward, GradientRegistry registry) {
  return [forward_model = std::move(forward),
          grad_registry = std::move(registry)](
             AbstractContext* ctx,
             absl::Span<AbstractTensorHandle* const> inputs,
             absl::Span<AbstractTensorHandle*> outputs) -> Status {

  return ops::DivNoNan(ctx, inputs[0], inputs[1], &outputs[0], "DivNoNan");
}

class CppGradients
    : public ::testing::TestWithParam<std::tuple<const char*, bool, bool>> {
 protected:
  void SetUp() override {
    TF_StatusPtr status(TF_NewStatus());
    TF_SetTracingImplementation(std::get<0>(GetParam()), status.get());
    status_ = StatusFromTF_Status(status.get());
    ASSERT_EQ(errors::OK, status_.code()) << status_.message();

                                    const std::vector<float>& expected_values) {
  std::unique_ptr<TF_Status, decltype(&TF_DeleteStatus)> status(
      TF_NewStatus(), TF_DeleteStatus);
  TF_Tensor* t = TFE_TensorHandleResolve(handle, status.get());
  ASSERT_EQ(TF_OK, TF_GetCode(status.get())) << TF_Message(status.get());
  std::unique_ptr<float[]> actual_values(new float[expected_values.size()]);