typedef typename RealType<T>::type RealT;
      Tensor t(DataTypeToEnum<RealT>::v(), shape);
      if (t.NumElements() != static_cast<int64_t>(v.size())) {
        status = absl::InvalidArgumentError(absl::StrCat(
            "Cannot construct a tensor with ", t.NumElements(),
            " from an initializer list with ", v.size(), " elements"));
        return;
      }

  if (actual.NumElements() != expected.NumElements()) {
    if (listener->IsInterested()) {
      *listener << "\nwas looking for tensor with " << expected.NumElements()
                << " elements, found tensor with " << actual.NumElements()
                << " elements";
      return false;
    }
  }

  for (int64_t i = 0, e = actual.NumElements(); i < e; i++) {

                                              StatusCallback done) {
  tsl::profiler::TraceMe traceme("PjRtDeviceContext::CopyDeviceTensorToCPU");
  if (device_tensor->NumElements() == 0) {
    VLOG(2) << "CopyDeviceTensorToCPU empty tensor";
    done(absl::OkStatus());
    return;
  }
  auto literal = std::make_unique<xla::MutableBorrowingLiteral>();

  ClientSession session(scope_test_);
  std::vector<Tensor> grad_result;
  TF_EXPECT_OK(session.Run({{x, {3.0f}}}, grad_outputs, &grad_result));
  EXPECT_EQ(grad_result.size(), 1);
  EXPECT_EQ(grad_result[0].NumElements(), 1);
  EXPECT_EQ(grad_result[0].flat<float>()(0), 17502.0f);
}

TEST_F(GradientsTest, MultiOutputNodeDependentOutputs) {
  auto x = Placeholder(scope_test_, DT_FLOAT);
  auto y0 = Square(scope_test_, x);

		if invalidLength(offset, t.length, len(bytes)) {
			err = SyntaxError{"truncated sequence"}
			return
		}
		offset += t.length
		numElements++
	}
	ret = reflect.MakeSlice(sliceType, numElements, numElements)
	params := fieldParameters{}
	offset := 0
	for i := 0; i < numElements; i++ {
		offset, err = parseField(ret.Index(i), bytes, offset, params)
		if err != nil {
			return
		}
	}
	return
}

};

std::vector<int64_t> IntTensorAsVector(const Tensor& t) {
  DCHECK(t.dtype() == DT_INT32 || t.dtype() == DT_INT64);
  std::vector<int64_t> result;
  result.reserve(t.NumElements());
  for (int i = 0; i < t.NumElements(); i++) {
    int64_t element = t.dtype() == DT_INT32
                          ? static_cast<int64_t>(t.flat<int32>()(i))
                          : t.flat<int64_t>()(i);
    result.push_back(element);

		{Group: "flowcontrol.apiserver.k8s.io", Version: "v1", Kind: "PriorityLevelConfigurationList"}:             {},
	}

	f := fuzz.New().NilChance(.5).NumElements(1, 1).RandSource(rand.NewSource(1))
	f.Funcs(
		func(s *runtime.RawExtension, c fuzz.Continue) {},
		func(s *metav1.LabelSelector, c fuzz.Continue) {
			c.FuzzNoCustom(s)

    // a reasonable amount apart
    T cur = 0;
    for (size_t i = 0; i < tensor->NumElements(); i++) {
      tensor_flat(i) = cur;
      cur += 5e-2;
    }
    // Fischer-Yates shuffle the array
    for (size_t i = tensor->NumElements() - 1; i >= 1; i--) {
      // j <- random integer 0 <= j <= i
      size_t j = random::New64() % (i + 1);
      // swap values at i, j

 public:
  explicit TapeVSpace(AbstractContext* ctx) : ctx_(ctx) {}
  ~TapeVSpace() override {}

  // Returns the number of elements in the gradient tensor.
  int64_t NumElements(AbstractTensorHandle* tensor) const override;

  // Consumes references to the tensors in the gradient_tensors list and returns
  // a tensor with the result.
  AbstractTensorHandle* AggregateGradients(

    t.reset(imm_t->Resolve(&status));
    if (!status.ok()) {
      return status;
    }
  }
  if (data.size() != t->NumElements()) {
    return tensorflow::errors::InvalidArgument(absl::StrCat(
        "Mismatched number of elements: \n", "Expected: ", data.size(), "\n",
        "Actual: ", t->NumElements(), "\n"));
  }
  memcpy(data.data(), t->Data(), t->ByteSize());
  return ::tensorflow::OkStatus();
}