TensorHandlePtr three_vector =
      VectorFloatTensorHandle({5., 6., 7.}, status.get());
  ASSERT_TRUE(TF_GetCode(status.get()) == TF_OK) << TF_Message(status.get());

  std::vector<TensorHandlePtr> vector_handles;
  vector_handles.reserve(2);
  vector_handles.push_back(std::move(two_vector));
  vector_handles.push_back(std::move(three_vector));
  std::unique_ptr<ParallelTensor> unknown_length_vector =

  if (!status->status.ok()) return;

  // Initialize a input_tensor vector with `nullptr` values.
  std::vector<const Tensor*> input_tensors_vector(num_inputs, nullptr);
  // A vector to keep track of newly created `tf::Tensor` objects.
  std::vector<Tensor> all_input_tensors;
  // Update the vector with information from `input_tensors` if provided.
  if (input_tensors != nullptr) {

#include "tensorflow/c/tf_status_helper.h"
#include "tensorflow/core/platform/errors.h"
#include "tensorflow/core/platform/test.h"

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());

    Model model, Model grad_model, AbstractContext* ctx,
    absl::Span<AbstractTensorHandle* const> inputs, bool use_function,
    double abs_error) {
  auto num_inputs = inputs.size();
  std::vector<AbstractTensorHandle*> outputs(num_inputs);
  auto s = RunModel(grad_model, ctx, inputs, absl::MakeSpan(outputs),
                    /*use_function=*/use_function);
  ASSERT_EQ(errors::OK, s.code()) << s.message();

See the License for the specific language governing permissions and
limitations under the License.
==============================================================================*/

#include <vector>

#include "tensorflow/c/eager/abstract_context.h"
#include "tensorflow/c/eager/abstract_tensor_handle.h"
#include "tensorflow/c/eager/c_api.h"
#include "tensorflow/c/eager/c_api_unified_experimental.h"

      return ops::MatMul(ctx, inputs[0], inputs[1], &outputs[0], transpose_a,
                         transpose_b, "MatMul");
    };
    Model MatMulGradModel = BuildGradModel(MatMulModel, registry_);
    std::vector<AbstractTensorHandle*> outputs(2);
    status_ =
        RunModel(MatMulGradModel, immediate_execution_ctx_.get(),
                 {A.get(), B.get()}, absl::MakeSpan(outputs), UseFunction());

  job_def->mutable_tasks()->at(task_index) =
      tensorflow::strings::StrCat("localhost:", port);
}

void CheckTFE_TensorHandleHasFloats(TFE_TensorHandle* handle,
                                    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());

build:release_arm64_linux --config=linux_arm64
build:release_arm64_linux --crosstool_top="@ml2014_clang_aarch64_config_aarch64//crosstool:toolchain"
build:release_arm64_linux --config=mkl_aarch64_threadpool
build:release_arm64_linux --copt=-flax-vector-conversions
test:release_arm64_linux --flaky_test_attempts=3

# The old gcc linux build options are preserved in the unsupported_*_linux
# configs. If your project fails to build with Clang, you can use these

  Status Shape(const std::vector<int64_t>** shape) const;
  TF_DataType dtype() const { return dtype_; }

  // Sets its output argument to a summary of the values of this tensor on every
  // component device.
  Status SummarizeValue(std::string& summary);

  std::vector<TensorHandlePtr> release_tensors() { return std::move(tensors_); }

  std::vector<TFE_TensorHandle*> tensors() const {

  }
}

absl::optional<std::vector<std::unique_ptr<ParallelTensor>>>
ParallelDevice::Join(
    const std::vector<PartialTensorShape>& expected_output_shapes,
    TF_Status* status) const {
  absl::optional<std::vector<std::unique_ptr<ParallelTensor>>> result;
  // Compute per-device per-output tensors
  std::vector<std::vector<TensorHandlePtr>> per_device_output_tensors;