EXPECT_EQ(22, product[3]);
  TF_DeleteStatus(status);
}
TEST(CAPI, Executor_MatMul_CPU) { Executor_MatMul_CPU(false); }
TEST(CAPI, Executor_MatMul_CPUAsync) { Executor_MatMul_CPU(true); }

void Deleter(void* data, size_t unused, void* tensor_handle) {
  TFE_DeleteTensorHandle(static_cast<TFE_TensorHandle*>(tensor_handle));
}

TEST(CAPI, TensorHandleOnDeviceMemory) {
  TF_Status* status = TF_NewStatus();

_initializers = _KerasLazyLoader(
    globals(), submodule="initializers", name="initializers")
setattr(_current_module, "losses", _losses)
setattr(_current_module, "metrics", _metrics)
setattr(_current_module, "optimizers", _optimizers)
setattr(_current_module, "initializers", _initializers)


# Do an eager load for Keras' code so that any function/method that needs to

  }
  return valid;
}
}  // namespace

void TFE_CallDLManagedTensorDeleter(void* dlm_ptr) {
  DLManagedTensor* dlMTensor = static_cast<DLManagedTensor*>(dlm_ptr);
  if (dlMTensor->deleter != nullptr) {
    dlMTensor->deleter(dlMTensor);
  }
}

void* TFE_HandleToDLPack(TFE_TensorHandle* h, TF_Status* status) {
  auto tf_dlm_context = GetDlContext(h, status);
  if (!status->status.ok()) {
    return nullptr;

    }                                                                        \
  }
DEFINE_GETATTR(TF_OperationGetAttrInt, int64_t, int64_t, i);
DEFINE_GETATTR(TF_OperationGetAttrFloat, float, float, f);
DEFINE_GETATTR(TF_OperationGetAttrBool, unsigned char, bool, b);
DEFINE_GETATTR(TF_OperationGetAttrType, TF_DataType, DataType, type);
#undef DEFINE_GETATTR

void TF_OperationGetAttrShape(TF_Operation* oper, const char* attr_name,

  import tensorflow_io_gcs_filesystem as _tensorflow_io_gcs_filesystem

# Lazy-load Keras v1.
_tf_uses_legacy_keras = (
    _os.environ.get("TF_USE_LEGACY_KERAS", None) in ("true", "True", "1"))
setattr(_current_module, "keras", _KerasLazyLoader(globals(), mode="v1"))
_module_dir = _module_util.get_parent_dir_for_name("keras._tf_keras.keras")
_current_module.__path__ = [_module_dir] + _current_module.__path__

                                                              TF_Status*);

// Creates a new TensorHandle from memory residing in the physical device
// device_name. Takes ownership of the memory, and will call deleter to release
// it after TF no longer needs it or in case of error.
//
// Custom devices must use TFE_NewCustomDeviceTensorHandle instead.
TF_CAPI_EXPORT extern TFE_TensorHandle* TFE_NewTensorHandleFromDeviceMemory(

  ASSERT_EQ(TF_OK, TF_GetCode(s_)) << TF_Message(s_);

  // Get attr
  AttrValue read_attr;
  GetAttr("test_attr_name", &read_attr);
  ASSERT_EQ(attr.DebugString(), read_attr.DebugString());

  // Retrieve the same attr after save/restore
  Reincarnate();
  AttrValue read_attr2;
  GetAttr("test_attr_name", &read_attr2);
  ASSERT_EQ(attr.DebugString(), read_attr2.DebugString());
}

        which would previously raise an error. This will correspond to an
        attribute name with an embedded '.' symbol (e.g. 'a.b'), which can only
        be accessed indirectly (e.g. through getattr and setattr). To set this
        up the user will first need to explicitly add the variable to the hparam
        object (e.g. "hparams.add_hparam(name='a.b', value=0.0)").
    *   Benchmark for tf.scan in graph and eager modes.

                                      absl::MakeSpan(in_gradients));
      if (!persistent_) {
        trace.backward_function_deleter(trace.backward_function);
      }
      if (!s.ok()) {
        return s;
      }
    } else {
      if (!persistent_) {
        trace.backward_function_deleter(trace.backward_function);
      }
      for (Gradient* grad : out_gradients) {
        if (grad != nullptr) {

  // device.
  std::function<void(Variable*)> variable_deleter = [&](Variable* to_delete) {
    to_delete->Destroy(context, status.get());
    ASSERT_TRUE(TF_GetCode(status.get()) == TF_OK) << TF_Message(status.get());
    delete to_delete;
  };
  std::unique_ptr<Variable, decltype(variable_deleter)> variable(
      Variable::Create(context, TF_FLOAT, /* Scalar */ {}, 0, device_name,