tensorflow::Tensor tensor(allocator.get(), DT_FLOAT, {1});

  EXPECT_THAT(
      GetPjRtCBufferFromTensor(&tensor),
      StatusIs(error::INTERNAL, HasSubstr(absl::StrCat(
                                    "Input tensor does not have PjRtBuffer"))));
}

TEST(TensorPjRtBufferUtilTest, GetPjRtCBufferFromTensorIncoorectType) {
  auto allocator = std::make_unique<AsyncValueAllocator>();

      const std::vector<TF_Tensor*>& input_tensors,
      const absl::optional<std::vector<int64_t>>& expected_shape) {
    // Create input_shapes.
    TF_ShapeAndTypeList* input_shapes =
        TF_NewShapeAndTypeList(input_shapes_vec.size());
    for (size_t i = 0; i < input_shapes_vec.size(); ++i) {
      const auto& input_shape = input_shapes_vec[i];

namespace gradients {
namespace {

static Status SafeConj(AbstractContext* ctx, AbstractTensorHandle* const input,
                       AbstractTensorHandle** output, const char* name) {
  auto dtype = input->DataType();
  if (DataTypeIsFloating(BaseType(dtype)) ||
      DataTypeIsInteger(BaseType(dtype))) {
    return tensorflow::ops::Identity(ctx, input, output, name);
  } else if (!DataTypeIsComplex(BaseType(dtype)) &&

      absl::StrCat("    signature {"
                   "      name: 'MatMulFunction'"
                   "      input_arg {"
                   "        name: 'a'"
                   "        type: DT_FLOAT"
                   "      }"
                   "      input_arg {"
                   "        name: 'b'"
                   "        type: DT_FLOAT"
                   "      }"

                             std::vector<AbstractTensorHandle*>* params) {
  tracing::TracingTensorHandle* handle = nullptr;
  for (auto input : inputs) {
    PartialTensorShape shape;
    TF_RETURN_IF_ERROR(input->Shape(&shape));
    TF_RETURN_IF_ERROR(dyn_cast<tracing::TracingContext>(ctx)->AddParameter(
        input->DataType(), shape, &handle));
    params->emplace_back(handle);
  }
  return absl::OkStatus();
}

    device_name_ = name;
    return absl::OkStatus();
  }

  Status AddInput(AbstractTensorHandle* input) override {
    GraphTensor* t = dyn_cast<GraphTensor>(input);
    if (!t) {
      return tensorflow::errors::InvalidArgument(
          "Unable to cast input to GraphTensor");
    }
    TF_AddInput(op_.get(), t->output_);
    return absl::OkStatus();
  }

}

// Specification for an expected edge.
// src is either:
// - input name (as it appears in FunctionDef)
// - name of output tensor (in nested "add:z:0" format)
// dst is either:
// - output name (as it appears in FunctionDef)
// - <name_of_node>:<index_of_this_input_into_node> (this looks the same as
//      output tensor naming, but it the index is actually an input index)
struct EdgeSpec : public std::pair<string, string> {

    Model model, AbstractContext* ctx,
    absl::Span<AbstractTensorHandle* const> inputs, int input_index,
    float* expected_grad, int num_grad, bool use_function,
    double abs_error = 1e-2) {
  Status s;
  AbstractTensorHandlePtr numerical_grad;
  {
    AbstractTensorHandle* numerical_grad_raw;
    s = CalcNumericalGrad(ctx, model, inputs, input_index, use_function,
                          &numerical_grad_raw);

  // Build an abstract input tensor.
  TFE_Context* eager_ctx =
      TF_ExecutionContextGetTFEContext(eager_execution_ctx, status.get());
  ASSERT_EQ(TF_OK, TF_GetCode(status.get())) << TF_Message(status.get());
  TFE_TensorHandle* input_eager = TestScalarTensorHandle(eager_ctx, 2.0f);
  TF_AbstractTensor* input_t =
      TF_CreateAbstractTensorFromEagerTensor(input_eager, status.get());

}
Status TapeOperation::AddInput(AbstractTensorHandle* input) {
  TF_RETURN_IF_ERROR(parent_op_->AddInput(input));
  forward_op_.inputs.push_back(input);
  return OkStatus();
}
Status TapeOperation::AddInputList(
    absl::Span<AbstractTensorHandle* const> inputs) {
  TF_RETURN_IF_ERROR(parent_op_->AddInputList(inputs));
  for (auto input : inputs) {
    forward_op_.inputs.push_back(input);
  }
  return OkStatus();
}