: inputs_(inputs.begin(), inputs.end()) {}

  InputList(const absl::Span<const Input>& inputs)  // NOLINT(runtime/explicit)
      : inputs_(inputs.begin(), inputs.end()) {}

  InputList(
      const std::initializer_list<Output>& out) {  // NOLINT(runtime/explicit)
    for (auto const& x : out) {
      inputs_.push_back(x);
    }
  }

}

Operation::Inputs Operation::GetInputs(Node* node) {
  Operation::Inputs inputs;
  if (node != nullptr) {
    inputs.resize(node->num_inputs(), {nullptr, -1});
    for (const Edge* e : node->in_edges()) {
      if (e->IsControlEdge()) continue;
      inputs[e->dst_input()] = std::make_pair(e->src(), e->src_output());
    }
  }
  return inputs;
}

Input::Initializer::Initializer(

    for (int i = 0; i < num_inputs; ++i) {
      inputs_.push_back(ops::Placeholder(scope_, dtype));
    }
  }

  void CreateLoop(const ops::CondGraphBuilderFn& cond,
                  const ops::BodyGraphBuilderFn& body,
                  const std::vector<Output>* inputs = nullptr) {
    if (inputs == nullptr) inputs = &inputs_;
    TF_ASSERT_OK(ops::BuildWhileLoop(scope_, *inputs, cond, body, "test_loop",

    return errors::InvalidArgument(
        "Must specify a gradient input for each output.");
  }
  std::vector<bool> reachable_nodes = GetReachableNodes();
  for (const Output& input : inputs_) {
    if (!reachable_nodes[input.node()->id()]) {
      return errors::InvalidArgument(
          "Cannot compute the partial derivative for node '",
          input.node()->name(),
          "' as it's unreachable from the output node(s).");

  }

  template <typename T>
  void Run(const std::vector<Input::Initializer>& input_values,
           const std::vector<T>& expected_output_values) {
    ClientSession session(scope_);

    DCHECK_EQ(input_values.size(), inputs_.size());
    ClientSession::FeedType feeds;
    for (int i = 0; i < inputs_.size(); ++i) {
      feeds.emplace(inputs_[i], input_values[i]);
    }

    std::vector<Tensor> out_tensors;

  Init(2);
  // Create nested loop:
  //  while (input1 < 6) {
  //    inner_input1 = input1
  //    while (inner_input1 < 3) {
  //      input2 += 1
  //      inner_input1 += 2
  //    }
  //    input1 += input2
  //  }
  //
  // Expected execution with initial values input1 = input2 = 0:
  //
  // outer inner               inner_
  // step# step# input1 input2 input1
  // ------------------------------------

  }

  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();
  }
  Status AddInputList(absl::Span<AbstractTensorHandle* const> inputs) override {

  EXPECT_EQ(TF_OK, TF_GetCode(s)) << TF_Message(s);
  TF_DeleteStatus(s);
}

void CSession::SetInputs(
    std::vector<std::pair<TF_Operation*, TF_Tensor*>> inputs) {
  DeleteInputValues();
  inputs_.clear();
  for (const auto& p : inputs) {
    inputs_.emplace_back(TF_Output{p.first, 0});
    input_values_.emplace_back(p.second);
  }
}

void CSession::SetOutputs(std::initializer_list<TF_Operation*> outputs) {

  std::vector<const Tensor*> inputs;
  inputs.push_back(a);
  inputs.push_back(b);
  TF_ASSERT_OK_AND_ASSIGN(auto execute_outputs,
                          RunExecutable(inputs, {}, result, executable));

  TF_EXPECT_OK(PopulateCtxOutputsFromPjRtExecutableOutputs(
      /*num_missing_prefix_ctx_inputs=*/0, inputs, {}, *result,

TF_Operation* RandomUniform(TF_Operation* shape, TF_DataType dtype,
                            TF_Graph* graph, TF_Status* s);

// Split `input` along the first dimension into 3 tensors
TF_Operation* Split3(TF_Operation* input, TF_Graph* graph, TF_Status* s,
                     const char* name = "split3");

bool IsPlaceholder(const tensorflow::NodeDef& node_def);